Chapter Ten - Some Casting Projects

 Before acquiring my home foundry, I spent many years watching other hobbyists and machinists make patterns, cast parts, and solve engineering problems by producing their own metal components. The ability to create a needed part from raw metal seemed almost magical. Today, affordable tools, readily available information, and small electric furnaces have made metal casting accessible to hobbyists working in home workshops.

\section{Pattern Making}

Every casting begins with a pattern. The pattern is a physical model of the desired part and is used to create the cavity in the mold that will later be filled with molten metal. Patterns can be made from wood, plastic, metal, or foam. One of the enjoyable aspects of foundry work is that it combines many different skills, including woodworking, machining, fabrication, and drafting. \index{pattern}

In recent years I have found that three-dimensional printing is one of the easiest methods of producing patterns. A pattern can be designed in CAD software, printed on a desktop 3D printer, sanded smooth, painted, and then used directly in the molding process. \index{printing, 3D} \index{pattern!printed}

Some castings require internal cavities or passages that cannot be formed by the pattern alone. In these cases, a core is used to create the negative space within the mold. Cores are typically made from sand mixed with a binder and are placed inside the mold before pouring. \index{pattern|core}

Small fillets and radii can be added to a pattern using wax, automotive body filler, or similar materials. These features improve mold filling and reduce stress concentrations in the finished casting. \index{fillets}

Depending on the molding process, surprisingly fine details can be reproduced in a casting. Surface textures, layer lines from 3D-printed patterns, engraved lettering, and small geometric features may all appear in the finished part.

\section{Lost Foam Casting}

For one-off projects and prototype work, a pattern may be constructed from polystyrene foam that has been cut, shaped, and glued together. The foam pattern is coated with a refractory slurry and allowed to dry. The coated pattern is then buried in loose sand without the use of a conventional mold. \index{lost foam}

During pouring, the molten metal vaporizes the foam and occupies the space previously filled by the pattern. This process is known as lost foam casting. While the method is simple and requires minimal tooling, it produces significant smoke and generally yields a rougher surface finish than conventional sand casting.

\section{Casting Shrinkage}

The pattern should have the same general shape as the finished part, but it must usually be made slightly larger to compensate for metal shrinkage during solidification and cooling. As the casting cools from the pouring temperature to room temperature, its dimensions decrease. \index{shrinkage allowance}

For many aluminum castings, a shrinkage allowance of approximately 1/8 inch per foot of length is adequate. Commercial foundries often use shrink rules or calculate shrinkage allowances based on the alloy being cast.

In my own projects, dimensional tolerances have rarely been critical, and shrinkage has not significantly affected the usefulness of the finished parts. However, shrinkage should always be considered when producing castings that must fit existing components or meet specific dimensional requirements.

\section{Casting Ingots}

Before you can cast parts, you will need a supply of metal. Suitable sources of aluminum include discarded lawn mower engines, vacuum pump housings, cookware, transmission housings, and other cast aluminum components. In general, if a part was originally produced by casting, it will often provide a suitable alloy for future casting projects. \index{aluminum!ingots}

New casters are frequently tempted to melt aluminum beverage cans and other thin packaging materials. While this can be done, it is generally inefficient. Because the material is thin and has a large surface area, much of the aluminum is lost to oxidation during melting. Large quantities of cans are required to produce a relatively small amount of usable metal. \index{aluminum!beverage cans}

I prefer to process dirty, oily, painted, or otherwise contaminated scrap in my propane furnace. After the metal has melted and impurities have been removed, the clean aluminum is poured into ingot molds for storage and future use. I commonly use a cast-iron muffin pan to produce small ingots and a steel angle-iron mold to produce triangular-section ingots. \index{ingot mold}

These clean ingots become the feedstock for my electric resistance furnace. By charging the electric furnace with previously processed metal, I reduce contamination, minimize smoke and fumes, and improve the overall casting experience. In practice, the propane furnace serves as a metal-processing furnace, while the electric resistance furnace serves as a clean melting and casting furnace.

\section{Mill Pendant and Wrench Hanger}

While working in the machine shop, I identified a need for a hanger to hold the milling machine control pendant and the vise wrench. I designed the required part in Onshape and produced the first version using a desktop 3D printer and polylactic acid (PLA) filament. \index{PLA filament}

The plastic part performed well in service for a period of time, but it was eventually damaged and broken. Rather than printing another plastic replacement, I decided to use the design as the basis for a permanent metal part.

A second PLA model was printed and used as the pattern for a sand casting. The pattern was rammed up in oil-bonded sand, removed from the mold, and replaced by molten aluminum. After the casting had cooled, it was removed from the mold, the sprue was cut off, the back surface was sanded flat, and the mounting holes were drilled to final size.

The finished aluminum hanger is considerably stronger than the original plastic version and has proven to be well suited for shop use. I expect the machine shop will continue using this part for many years.

\section{Metal Carbide Part Core}

Another opportunity to apply metal casting arose at work during an experiment involving the ramming of metal carbide powder into a mold that combined features from two separate parts. I believed the design could be improved by incorporating an aluminum core with rounded corners. The rounded edges would form fillets in the compacted powder part, reducing stress concentrations and improving the likelihood of successful processing. \index{fillets}

The core was designed in CAD and printed on a 3D printer using polylactic acid (PLA) filament. Because of its large size, the pattern was printed with only 5\% infill to reduce material consumption and printing time. Care was taken to provide sufficient draft so the core could be easily removed from the molded carbide part. After printing, the pattern was sanded smooth and coated with several layers of filling primer. \index{PLA filament} \index{draft}

The core was cast in an oil-bonded sand mold and required approximately 12 kilograms of aluminum, making it the largest casting I had produced at that time. After cooling, the casting was removed from the mold, the sprue was cut off, and it was finished by sanding and smoothing the surfaces.

The casting contained several surface pits and voids that were not intended. I suspect these defects were caused by entrapped air, turbulence during pouring, or sand inclusions in the mold. Large castings may require a riser, which is a reservoir of molten metal connected to the casting. As the casting cools and shrinks, the riser supplies additional metal to compensate for solidification shrinkage and reduce the formation of internal voids.

Although the casting was not cosmetically perfect, it was fully functional for its intended purpose. \index{casting defects}

The completed core was taken to work and used in the production experiment. This project demonstrated how a home foundry can be used to rapidly produce specialized tooling and fixtures that would otherwise be difficult or expensive to obtain.

\section{Cast Parts for Furnace Mast Support}

During the construction of the furnace described in this book, I needed specialized brackets to attach the lid-lifting mast to both the furnace body and the removable lid. Commercial hardware such as antenna mounting brackets was considered, but the available parts did not fit the application particularly well. Because a home foundry was available, designing and casting custom components became a practical alternative.

The brackets were designed in Onshape and the patterns were produced using a 3D printer and polylactic acid (PLA) filament. The patterns were rammed up in oil-bonded sand molds and cast in aluminum. \index{PLA filament} \index{OnShape}

After cooling, the castings were removed from the molds, the sprues were cut away, and the surfaces were sanded smooth. The mounting holes were then drilled to final size and the parts were painted with high-temperature black paint.

The finished brackets were used to attach the mast assembly to the furnace. This project illustrates one of the greatest advantages of having a home foundry: when a specialized component is needed, it is often possible to design and manufacture the part rather than search for a commercial substitute.

\section{Shop-Built Foundry Tools}

As my foundry activities expanded, I accumulated crucibles of various sizes and discovered the need for specialized handling equipment. Commercial foundry tools are available, but they can be expensive and are not always suited to a particular crucible size. As a result, I fabricated my own lifting tongs and pouring shanks from rectangular and round steel stock, assembling the tools by welding. \index{tongs}

Additional molding flasks were constructed as needed from dimensional lumber and common hardware. Building these flasks allowed me to produce larger castings without purchasing commercial foundry equipment. \index{molding!flasks}

Like David Gingery before me, I learned to repurpose ordinary items for foundry service. Slotted spoons purchased from a thrift store became skimmers for removing dross from molten metal. A flour sifter was adapted for screening molding sand. Cast-iron cookware found new life as ingot molds. \index{ingot mold} \index{sand riddle} \index{Gingery, David}

One of the lessons learned from building a home foundry is that the furnace itself is only the beginning. As new projects arise, additional tools, fixtures, molds, and handling equipment are often needed. Many of these items can be fabricated in the home workshop, further expanding the capabilities of the foundry and the skills of its operator.

\section{Lessons Learned}

\textbf{Most castings succeed on the second or third attempt} rather than the first. Defects such as misruns, shrinkage cavities, gas porosity, and sand inclusions are common, particularly when learning a new process. Success comes from carefully observing the results, changing one variable at a time, and persistently refining the mold and pouring technique. If you are willing to learn from your mistakes, each failed casting becomes a valuable lesson. \index{casting defects}

\textbf{Draft and fillets matter.} Without adequate draft, removing the pattern from the mold can be difficult and may damage the mold cavity. Fillets improve mold filling, reduce stress concentrations, and make patterns easier to withdraw from the sand. These features should be considered from the earliest stages of the design process. \index{draft} \index{fillets}

\textbf{Pattern quality directly affects casting quality.} Casting processes are capable of reproducing surprisingly fine details, including surface textures, tool marks, and layer lines from 3D-printed patterns. Time spent improving the pattern will generally be rewarded with better surface finish, more accurate castings, and fewer molding problems.

\textbf{A home foundry provides a level of independence} that is difficult to achieve by other means. Instead of searching for a commercial solution, the foundry operator can often design and manufacture a custom part, tool, or fixture as needed. The ability to transform an idea into a functional object is both practical and deeply satisfying.

\section{\;\,In Closing}

The project described in this book provides more than a means of melting metal. Along the way, the builder gains experience in design, fabrication, electrical wiring, temperature control, refractory construction, and furnace operation. These skills can be applied to countless future projects and often prove to be as valuable as the castings themselves.

At the conclusion of the project, the builder is left with something equally valuable: a practical tool capable of supporting future work in metal casting, heat treatment, materials experimentation, and shop fabrication. The furnace is not the end of the journey, but rather the beginning of a new set of possibilities.

Whether the next project is a replacement machine part, a custom tool, an artistic casting, or an entirely new machine, the furnace provides the capability to transform ideas into physical objects. In that sense, the greatest value of the project is not the furnace itself, but the opportunities it creates.

If you have questions, comments, suggestions, or would simply like to share your own experiences, please feel free to contact me at vincethigpen@yahoo.com. I try to check my email regularly and would enjoy hearing about your personal metal casting journey.

Thank you for reading this book, and best wishes for success in your future projects.

Chapter Nine - Evaluating Performance

 Before beginning this chapter, it is worth noting that it contains more equations and calculations than any previous chapter in this book. This is not intended to be a textbook on thermodynamics or heat transfer. Rather, the measurements, calculations, and performance data presented here are used to evaluate how effectively the furnace performs its intended functions.

The goal of this chapter is to answer practical questions that builders and operators are likely to have. How quickly does the furnace heat up? How much electrical energy does it consume? How much heat is lost through the insulation? How efficiently does it melt aluminum? The calculations presented here provide a framework for answering these questions and for comparing the performance of this furnace to other designs.

Readers who are primarily interested in building and operating the furnace may choose to skim some of the mathematical details. However, the results and conclusions drawn from these calculations provide valuable insight into the performance of the finished equipment.

% Imagine a chapter containing these graphs:

% Temperature vs. Time 

% Duty Cycle vs. Temperature 

% Power Consumption vs. Temperature 

% Temperature Gradient Through Furnace Wall 

% Shell Temperature vs. Hot-Zone Temperature 

% Cost of Operation vs. Temperature

\section{Test Equipment}

\section{Testing Equipment}

The following equipment was used to collect the performance data presented in this chapter.

\begin{description}

\item[PID Controller]

The Inkbird ITC-100VH PID controller was used to display and record furnace temperature through a Type K thermocouple installed in the lid of the heating chamber.  The controller also signals when power is being supplied to the heating elements.

\item[Infrared Thermometer]

An inexpensive handheld infrared thermometer was used to measure the temperatures of the rear firebrick, furnace shell, lid, and other accessible surfaces during operation.

\item[Digital Ammeter]

A panel-mounted digital ammeter and current transformer were used to monitor heating element current during furnace operation.

\item[Digital Scale]

An inexpensive digital kitchen scale was used to measure the mass of aluminum used during melting trials and to verify the mass of completed castings.

\item[Stopwatch]

A stopwatch was used to record heat-up times, melting times, and other time-dependent measurements.  Using the stopwatch was not very accurate and another method, discussed later in the chapter, was devised to capture duty cycle times.

\item[Type K Thermocouple]

A Type K thermocouple installed in the furnace lid provided temperature feedback to the PID controller and served as the primary temperature measurement device.

\item[Volt-Ohm-Milliammeter] A Fluke portable volt-ohm-milliammeter was used for troubleshooting and obtaining electrical circuit values.

\end{description}

The measurements presented in this chapter should be considered engineering estimates rather than laboratory-grade measurements. The primary objective of the testing was to characterize furnace performance and identify opportunities for future improvement rather than to produce highly precise scientific data.

% This establishes measurement credibility.

% \SI{240}{\volt} 

% \SI{16.7}{\ampere} 

% \SI{4000}{\watt}

% \SI{14.9}{\ohm}

% \SI{1000}{\celsius}

\section{Heat-Up Performance}

Measure:

Ambient temperature -- 25 \dg C \\

Time to 100 \dg C -- 7 minutes 20 seconds\\

Time to 200 \dg C -- 10 minutes 15 seconds\\

Time to 400 \dg C -- 17 minutes 5 seconds\\

Time to 600 \dg C -- 23 minutes 10 seconds\\

Time to 800 \dg C -- \\

Time to 1000 \dg C --

\begin{figure}[H] % a sample heat up plot

\centering

\begin{tikzpicture}

\begin{axis}[

    width=1.0\textwidth,

    height=0.5\textwidth,

    grid=major,

    minor tick num=1,

    xlabel={Time (minutes)},

    ylabel={Temperature (\dg C)},

    title={Furnace Heat-Up Test},

]

\addplot[

    color=red,

    very thick,

    mark=*,

]

table[

    x=time_min,

    y expr=(\thisrow{temperature_f}-32)*5/9,

    col sep=comma

]

{data/furnace_heatup.csv}; % sample time-temp data

\end{axis}

\end{tikzpicture}

\caption{Measured furnace temperature during heat-up.}

\label{fig:heatupcurve}

\end{figure}

\section{Electrical Power Consumption}

Power consumption measurements were taken after the furnace had reached steady-state operation and the heating elements were operating at full power.  Electric power consumption was measured using the installed voltage-current meter on the control panel.  

Electric power consumed is calculated using either Equation \ref{eq:power1} or Equation \ref{eq:power2}.  I elected to calculate the power both ways as a check.

\begin{equationbox}

\begin{equation}

P = I^2R

    \label{eq:power1}

\end{equation}

\begin{equation}

P = IV

   \label{eq:power2}

\end{equation}

\end{equationbox}

% This is where your \SI{14.9}{\ohm} element resistance becomes useful.

The furnace was designed to consume approximately 4000 watts.  Using Equation \ref{eq:power1} and using the previously measured resistance of 14.9 ohms for the heating elements when instantaneous current displayed was 16.4 amps gave me a furnace power of 4008 watts.  This is approximately 100\% of the design power level.

\begin{align}

P &= I^2R \\

  &= (16.4\,\text{A})^2(14.9\,\Omega) \\

  &= 4008\,\text{W}

\end{align}

Using Equation \ref{eq:power2} and using the displayed instantaneous current of 16.4 amps and the displayed voltage of 243 volts, I calculated that the furnace power was 3985 watts.  This is approximately 99\% of the design power level.

\begin{align}

P &= IV \\

  &= (16.4\,\text{A})(243\,\text{V}) \\

  &= 3985\,\text{W}

\end{align}

The two calculated power values differ by less than one percent, indicating good agreement between the measurements and calculations. Several factors may contribute to this small difference.

First, the resistance value used in Equation \ref{eq:power1} was measured at room temperature with the furnace de-energized. The electrical resistance of Kanthal heating elements increases slightly as their temperature increases. Consequently, the actual operating resistance of the heating elements is somewhat greater than the value used in the calculation.

Second, the installed voltage-current meter is an inexpensive consumer-grade instrument intended for monitoring rather than precision measurement. Small errors in the displayed voltage and current values are therefore expected.

Considering these factors, the agreement between the two calculated power values is excellent and confirms that the furnace operates very near its intended design power of 4000 watts.

\section{Exterior Surface Temperatures}

Using the non-contact thermometer, I first confirmed that the workshop ambient temperature was approximately 29 \dg C by taking spot measurements of a dozen surfaces in the shop and averaging the result.

After the furnace came to equilibrium with a 1000 \dg C cavity temperature, I began taking spot surface temperature measurements on the shell (top, middle, bottom) at four locations 90\dg \, apart, at the point where the lid meets the body, four points on top of the furnace lid, the back of the control cabinet, and the top of the control cabinet.

Temperatures were taken with the shop exterior door closed and the room fan turned off.  

\begin{table}[H]

\centering

\caption[Exterior Surface Temps]{Exterior Surface Temps at 1000 \dg C Cavity Temp}

\label{tab:surfacetemps}

\begin{tabular}{|l|c|}

\hline

\textbf{Location} & \textbf{Temperature (\dg C)} \\

\hline

Shell - Top (Location 1) & 35 \\

Shell - Middle (Location 1) & 31 \\

Shell - Bottom (Location 1) & 29 \\

\hline

Shell - Top (Location 2) & 36 \\

Shell - Middle (Location 2) & 32 \\

Shell - Bottom (Location 2) & 30 \\

\hline

Shell - Top (Location 3) & 35 \\

Shell - Middle (Location 3) & 32 \\

Shell - Bottom (Location 3) & 31 \\

\hline

Shell - Top (Location 4) & 34 \\

Shell - Middle (Location 4) & 33 \\

Shell - Bottom (Location 4) & 29 \\

\hline

Lid Surface (Point 1) & 37 \\

Lid Surface (Point 2) & 38 \\

Lid Surface (Point 3) & 39 \\

Lid Surface (Point 4) & 38 \\

\hline

Lid-to-Body Joint (Average) & 59 \\

\hline

Control Cabinet - Rear Surface & 31 \\

Control Cabinet - Top Surface & 33 \\

\hline

Ambient Workshop Temperature & 29 \\

\hline

\end{tabular}

\end{table}

The measured exterior temperatures demonstrate the effectiveness of the insulating firebrick and ceramic fiber blanket insulation. Even with a furnace cavity temperature of 1000 °C, the exterior shell remained cool enough to touch comfortably. The control cabinet temperatures remained near ambient, confirming that the electrical components were adequately isolated from the heat generated by the furnace.

The highest measured exterior temperature was 59 \dg C at the interface between the furnace body and the removable lid. This location was approximately 30 \dg C above the ambient workshop temperature and likely represents the largest source of heat loss from the furnace.

The measured temperatures compare favorably with those of Furnace Version 1.0 and indicate that the insulation system used in Furnace Version 2.0 is highly effective.

\section{A Few Words About Insulation}

One of the most effective ways to reduce heat loss is to increase the thickness of the insulation surrounding the hot zone. As a general rule, increasing insulation thickness reduces the rate of heat transfer through the wall. For simple conductive heat transfer, doubling the insulation thickness approximately halves the heat flow.

Consider a freezer with two inches of insulation through which 6000 BTU of heat enters each day. If the insulation thickness is increased to four inches, the heat transfer may be reduced to approximately 3000 BTU per day. Increasing the insulation thickness to eight inches may reduce the heat transfer further to approximately 1500 BTU per day.

See Figure \ref{fig:wallthickness} and notice that each additional doubling of insulation produces a smaller reduction in heat loss than the previous layer. Eventually a point is reached where the cost, weight, and space required for additional insulation are no longer justified by the reduction in heat transfer.

\begin{figure}[H] 

\centering

\begin{tikzpicture}

\begin{axis}[

    width=0.9\textwidth,

    height=0.5\textwidth,

    grid=major,

    minor tick num=1,

    xlabel={Insulation Thickness (inches)},

    ylabel={Heat Gain (BTU per day)},

    title={Insulation Thickness vs Heat Gain},

]

\addplot[

    color=red,

    very thick,

    mark=*,

]

table[

    x=thick,

    y=BTU_day,

    col sep=comma

]

{data/insulation_vs_conduction.csv};

\end{axis}

\end{tikzpicture}

\caption{Effect of increasing insulation thickness.}

\label{fig:wallthickness}

\end{figure}

Furnace Version 1.0 performed well with a thinner insulating wall. Because Furnace Version 2.0 incorporates approximately two additional inches of insulation, it is expected to exhibit lower heat loss, cooler exterior surface temperatures, and improved thermal efficiency.

\vspace{.5cm}

\begin{equationbox}

For flat walls the heat transfer can be approximated by:

\begin{equation}

q = \frac{k At \Delta T}{L} = \frac{kAt(T_{Hot} - T_{Cold})}{L}

\end{equation}

where:

\begin{compactitem}

\item[$q$] -- Heat flux (W or J/s) 

\item[$k$] -- thermal conductivity (W / (m $\cdot$ K)

\item[$A$] -- Area of wall ($m^2$)

\item[$t$] -- time duration (seconds)

\item[$\Delta T$] -- Temperature difference (\dg C)

\item[$L$] -- Wall thickness (m) 

\end{compactitem}

\end{equationbox}

Choosing an insulator with low thermal conductivity (k) reduces the wall thickness required to retain thermal energy. Because of their low thermal conductivities, insulating firebrick and ceramic fiber blanket were selected as the primary insulating materials for this furnace.

\begin{table}[H]

    \caption[k-Values]{K-Values of Common Materials}

    \label{tab:k-values}

    \begin{center}

    \begin{tabular}{l|c} \hline

        \textbf{Material}     &   \textbf{k (W/m·K)} \\

        \hline

    Copper          & ~400 \\

    Aluminum      & ~205 \\

    Carbon steel & ~50 \\

    Fireclay brick & ~1 \\

    IFB              & ~0.15 -- 0.30 \\

    Ceramic fiber blanket & ~0.05 -- 0.15 \\

    Ceramic fiber board     &   ~0.10 -- 0.25   \\

    Air              & ~0.026 \\

        \hline

    \end{tabular}

    \par\medskip\footnotesize

     Thermal conductivity varies with temperature, density, and manufacturer.

    \end{center}

\end{table}

Although air has the lowest thermal conductivity listed in Table \ref{tab:k-values}, large air spaces are generally poor insulators because natural convection currents can transfer heat. Ceramic fiber blanket and insulating firebrick are effective because they trap countless small pockets of air, greatly reducing convective heat transfer while also limiting heat conduction.

\section{Rear Firebrick Temperatures}

This is one of the most interesting tests because you installed thermocouples specifically for this purpose.

Measure:

Rear brick \#1 \\

Rear brick \#2 \\

Rear brick \#3 \\

Rear brick \#4

This reveals:

Thermal gradients \\

Hot spots \\

Uniformity \\

\section{Controller Duty Cycle}

Because the PID controller cycles power to the heating elements, the furnace does not necessarily consume full rated power during the entire melt. Measuring duty cycle allows the average electrical power to be estimated more accurately. Better measurements of current, voltage, duty cycle, melt time, and charge temperature will produce a more realistic estimate of furnace efficiency.

Initially, I attempted to measure the furnace duty cycle by observing the heating element indicator LED on the PID controller, recording the elapsed time with a stopwatch, and manually noting the results. This method proved to be inaccurate because many of the heating cycles were very short and difficult to record reliably.

After discussing the problem with the OpenAI ChatGPT 5.5 language model, it was suggested that I record a video of the PID controller while the furnace was operating. A cellular telephone mounted on a tripod was positioned so that the indicator LED and the was visible during the recording.

The video was later reviewed frame-by-frame. Each time the heating element indicator illuminated, the corresponding video timestamp was recorded. By summing the total time that the indicator was illuminated and dividing by the total elapsed test time, an approximate controller duty cycle was determined.

This method proved to be significantly more accurate than direct observation and provided a permanent record that could be reviewed if necessary.

Recorded:

Duty cycle at 400 \dg C -- 32\%\\

Duty cycle at 600 \dg C -- 41\%\\

Duty cycle at 800 \dg C -- 56\%\\

Duty cycle at 1000 \dg C -- 65\%

We can see that the heater has to work harder to maintain higher temperatures.

\section{Aluminum Melting Trials}

This section will likely be of greatest interest to many readers because melting metal is the primary reason for building an electric resistance shop furnace.

The aluminum melting trial serves two purposes. First, it demonstrates the furnace's ability to melt metal under normal operating conditions. Second, the measured melting time and power consumption data can be used later to estimate furnace efficiency.

\begin{cautionbox}

Before beginning any melting trial, place an empty ingot mold near the furnace. If a furnace malfunction occurs, the molten metal can be safely poured into the mold before it solidifies. \\

Do not allow molten aluminum to solidify inside a fireclay crucible. As aluminum freezes it expands and may crack the crucible. Even if no visible damage is observed, the crucible may have been weakened and could fail during a subsequent firing.

\end{cautionbox}

\begin{warningbox}

Plan the pour before beginning the melt. Wear all required personal protective equipment, including:  

\begin{multicols}{2}

\begin{compactitem}

    \item Face shield

    \item Heat-resistant gloves

    \item Long-sleeved cotton shirt

    \item Leather or rubber apron

    \item Long cotton or denim pants

    \item Safety shoes or boots

\end{compactitem}

\end{multicols}

Place spare firebricks or another suitable heat-resistant surface nearby before lifting the crucible. Position the mold on a non-flammable surface such as concrete or dry sand, and ensure that the area between the furnace and mold is free of obstacles and tripping hazards. \\

Practice the lifting and pouring motions before melting metal. Once the aluminum is molten, there is little time to decide where to walk, where to set the crucible down, or how the pour will be performed.

\end{warningbox}

To prepare for the test, weigh out 1 kilogram of clean aluminum and gather all required personal protective equipment, material handling tools, and molds.

Set up the furnace in an area free of combustible materials. Connect the furnace to the 240 VAC power supply, open the control cabinet, and close the 25 amp circuit breaker. The cooling fans should immediately begin operating. Once the circuit breaker is closed, components inside the control cabinet are energized. Close the cabinet door and leave it closed during operation.

Place a graphite saucer on the furnace floor and position the crucible on top of the saucer. Close the lid, set the PID controller setpoint to 750 \dg C, press the Start Pushbutton, and preheat the furnace and crucible.

When the furnace has reached operating temperature, load the 1 kilogram aluminum charge into the crucible and close the lid. Press the Start Pushbutton, and begin timing the melt.

Observe the furnace during operation and record any data required for later analysis, including heat-up time, melting time, shell temperatures, duty cycle, and power consumption. Once the aluminum has completely melted, proceed with the planned pour and record the total elapsed time.

Repeat the melt with 3 pounds and 5 pounds of aluminum.  These additional datasets will be useful when you calculate furnace efficiency later.

\section{Energy Consumption Per Melt}

Determine:

kWh consumed for a typical melt.

\begin{equationbox}

\begin{equation}

E_{\text{kWh}} =

\frac{P_{\text{furnace}}Dt}{1000}

\end{equation}

where: 

\begin{tabular}{>{$}l<{$}l}

P_{\text{furnace}} & -- furnace power when energized (kW) \\

D                  & -- duty cycle as a decimal \\

t                  & -- melt time (hours)

\end{tabular}

\end{equationbox}

Substituting values we measured and recorded:

\begin{align}

E_{\text{kWh}} &=

(4.0\ \text{kW})(0.60)\left(1\text{hr}\right) \\

&= 2.4\ \text{kWh}

\end{align}

Then calculate cost per melt using your local electricity rate.  Here at Apex, NC we are charged \$0.12 per kWh.   This calculated cost does not account for the energy used for pre-heating the furnace and crucible.

\begin{align}

\text{Cost} &= (2.4\ \text{kWh})(\$0.12/\text{kWh}) \\

&= \$0.29

\end{align}

It costs approximately \$0.30 per hour to operate this furnace.  This is very economical when you compare it to the alternatives available to hobbyists. 

\section{Furnace Efficiency Estimate}

To estimate the efficiency of the furnace, I conducted a series of runs at various setpoint temperatures to discover how widely the duty cycle varies during operation.  Next, I calculated the calorimetry of completely melting one kilogram of aluminum metal and giving the molten aluminum about 40 \dg C of superheat.

Calculate:

\begin{compactitem}

\item{Energy required to heat metallic aluminum}

\item{Energy required for fusion}

\item{Energy required to heat liquid aluminum}

\item{Electrical energy consumed}

\end{compactitem}

Once we know how much energy is required to melt 1 kilogram of aluminum, we actually melt 1 kilogram of aluminum and compare the amount of energy used.

\begin{equationbox}

\begin{equation}

Q = mc\Delta T = mc(T_{Hot} - T_{Cold})

\label{eq:heat1}

\end{equation}

where:

\begin{compactitem}

\item[$Q$] -- Heat (J)

\item[$m$] -- Mass (kg)

\item[$c$] -- Specific heat capacity (J/kg·\dg C)

\item[$\Delta T$] -- Temperature change (\dg C)

\end{compactitem}

\end{equationbox}

To melt the aluminum, we have to heat it from room temperature ($T_{Cold}$) to the melting point ($T_{Hot}$).  Assume that the room temperature is 20 °C.  Aluminum melts at 660 °C.   Assume we have 1 kilogram of aluminum (the mass, m) in the crucible.

The specific heat capacity of SOLID aluminum is 900 Joules/kilogram \dg C (c).  Use Equation \ref{eq:heat1} to calculate the heat required to bring the 1 kilogram of aluminum to the melting point ($Q_1$):  

\begin{align}

Q_1 &=

(1\,\text{kg})

\left(

900\,\frac{\text{J}}{\text{kg}\cdot{}^\circ\text{C}}

\right)

(660^\circ\text{C}-20^\circ\text{C}) \\

&=

(1\,\text{kg})

\left(

900\,\frac{\text{J}}{\text{kg}\cdot{}^\circ\text{C}}

\right)

(640^\circ\text{C}) \\

&= 576{,}000\ \text{J} \\

&= 576\ \text{kJ}

\end{align}

After the metal temperature reaches the melting point, additional heat must be added to cause the phase change from solid to liquid.  This heat ($Q_2$) is called the latent heat of fusion ($L_f$), and for aluminum it is 390 kiloJoules per kilogram. \index{latent heat}

\begin{equation}

Q_2 = mL_f

\label{eq:heat2}

\end{equation}

\begin{align}

Q_2 &=

(1\,\text{kg})

\left(

390{,}000\,\frac{\text{J}}{\text{kg}}

\right) \\

&= 390{,}000\ \text{J} \\

&= 390\ \text{kJ}

\end{align}

The total amount of heat needed to melt 1 kilogram of aluminum is the heat needed to raise its temperature to the melting point ($Q_1$), plus the latent heat of fusion ($Q_2$) needed to make it melt and change phase to a liquid.

\begin{equation}

Q_T = Q_1 + Q_2

\end{equation}

\begin{align}

Q_T &= 576\ \text{kJ} + 390\ \text{kJ} \\

&= 966\ \text{kJ}

\end{align}

If additional heat is supplied to the aluminum to raise its temperature above the melting point for pouring ($Q_3$), we can use Equation \ref{eq:heat1} again to find out how much heat is required.  The specific heat capacity of LIQUID aluminum is 1180 Joules/kilogram °C (c).  Assume we heated the aluminum to 700 °C.

\begin{equation}

Q_3 = mc(T_{Hot} - T_{Cold})

\end{equation}

\begin{align}

Q_3 &=

(1\,\text{kg})

\left(

1180\,\frac{\text{J}}{\text{kg}\cdot{}^\circ\text{C}}

\right)

(700^\circ\text{C}-660^\circ\text{C}) \\

&=

(1\,\text{kg})

\left(

1180\,\frac{\text{J}}{\text{kg}\cdot{}^\circ\text{C}}

\right)

(40^\circ\text{C}) \\

&= 47{,}200\ \text{J} \\

&= 47.2\ \text{kJ}

\end{align}

\begin{equationbox}

\begin{equation}

Q_T = Q_1 + Q_2 + Q_3\

    \label{eq:heat3}

\end{equation}

where:

\begin{compactitem}

\item[$Q_T$] -- Total amount of heat required. (J)

\item[$Q_1$] -- Heat to raise the metal to 660 \dg C. (J) -- Use Equation \ref{eq:heat1}

\item[$Q_2$] -- Heat to completely melt the charge. (J) -- Use Equation \ref{eq:heat2}

\item[$Q_3$] -- Heat to raise the liquid to 700 \dg C. (J) -- Use Equation \ref{eq:heat1}

\end{compactitem}

\end{equationbox}

To bring 1 kilogram of solid aluminum from room temperature to liquid at 700 \dg C, you add the heat required to raise the aluminum to the melting point ($Q_1$), plus the latent heat required to cause the phase change ($Q_2$), plus the heat added to the liquid to bring the temperature to 700 \dg C ($Q_3$).

Using Equation \ref{eq:heat3} and substituting our values:

\begin{align}

Q_T &= 576\ \text{kJ} + 390\ \text{kJ} + 47.2\ \text{kJ} \\

&= 1013.2\ \text{kJ} \\

&= 1.0132\ \text{MJ}

\end{align}

The first melt requires more energy than subsequent melts because the furnace materials and the crucible have to be heated up for that first melt.  It is more accurate to pre-heat the furnace and crucible, then add the mass of room temperature aluminum for the test.

If the furnace were 100\% efficient, all of the energy supplied to it would go into melting the aluminum.  The furnace supplies 4000 Joules per second.  966,000 Joules are required to completely melt 1 kilogram of aluminum.  It would take 966,000 Joules divided by 4000 Joules per second, or 242 seconds -- about 4 minutes, to melt 1 kilogram of aluminum.

Of course it took much more time to actually melt the metal.

The efficiency of the furnace ($\eta$) is the amount of energy required to melt the aluminum divided by the amount of energy supplied to the furnace multiplied by 100\%.  The actual amount of energy is found by multiplying the rated power of the furnace by the duty cycle and time (in seconds).  See Equation \ref{eq:eta_furnace} below.

\begin{equationbox}

\begin{equation}

\eta = \Bigg(\frac{Power Output}{Power Supplied}\Bigg) \times 100\%

    \label{eq:efficient1}

\end{equation}

\begin{equation}

\eta =

\left(

\frac{Q_{\text{required}}}

{P_{\text{furnace}} D t}

\right)

\times 100\%

\label{eq:eta_furnace}

\end{equation}

where:

\begin{tabular}{>{$}l<{$} l l}

Q_{\text{required}} & = & \text{heat required to melt and heat the aluminum, (J)} \\

P_{\text{furnace}}  & = & \text{furnace power when energized, (W)} \\

D                   & = & \text{duty cycle expressed as a decimal fraction} \\

t                   & = & \text{elapsed melt time, (s)}

\end{tabular}

\end{equationbox}

Assume it takes 40 minutes to completely melt 1 kilogram of aluminum and the furnace duty cycle is 60\%.

\begin{align}

\eta &=

\left(

\frac{966{,}000\ \text{J}}

{(4000\ \text{J/s})(0.60)(2400\ \text{s})}

\right)

\times 100\% \\

&= 16.8\%

\end{align}

I pre-heated the furnace and crucible, added 1 kilogram of room temperature aluminum to the crucible, and then timed how long it took for the charge to completely melt. I had to track the duty cycle of the heater to calculate the actual amount of energy supplied.  There is some uncertainty in this measurement, but it gives me an idea of how much of the supplied energy escapes or is wasted.

So, about 17\% efficient.  This is a rough measurement, but it shows that there is room for improvement.  One improvement that was identified during this testing was that heat appears to be leaking from the lid-to-body interface.  Improved sealing here seems to be needed.  

\section{Comparison With Propane Furnace}

At the time of this writing my propane furnace is 14 years old.  It is a very simple device and it continues to give me good service.  Because I have now operated the propane furnace and the electric resistance furnace, I can do a side-by-side comparison of both.

\begin{table}[H]

    \caption[Comparison]{Propane versus Electric Shop Furnace}

    \label{tab:compare}

    \begin{center}

    \begin{tabular}{|l|c|c|} \hline

        \textbf{Criterion}  & \textbf{Propane}  &   \textbf{Electric} \\

        \hline

    Noise          & Very loud  &  Silent \\

    Heat-up time     &  Rapid      &  Slower \\

    Operating cost & About \$4 per hour  &  About \$0.30 per hour \\

    Convenience  & Some setup  &  Plug-in and go  \\

    Indoor use       & Dirty - Outdoors only  &  Clean - Indoor use \\

    Temp Control     & Poor   &  Good \\

        \hline

    \end{tabular}

    \par\medskip\footnotesize

     The propane furnace continues to be useful for cleaning dirty scrap metal.

    \end{center}

\end{table}

The propane furnace must be used outdoors because the combustion of propane generates poisonous carbon monoxide.  Outdoor use cannot occur on rainy days due to the risk of a steam explosion.  The electric furnace generates no smoke or fumes and it can run indoors anytime.

The propane furnace makes a roaring sound during operation.  This generates unwanted attention from neighbors and curious neighborhood children.  The electric resistance furnace is completely silent and is unseen indoors.

The propane furnace heats up very quickly, but controlling the temperature of the charge is difficult.  I have boiled the zinc out of brass and zamak alloy because I was unable to limit and control the temperature.  The electric resistance furnace has very tight temperature control and this makes it useful for melting alloys and heat treatment of metals.

The propane used in the propane furnace costs about \$4.00 per hour of operation.  I use a 20-pound steel storage tank, and if I exhaust the tank I must drive to the hardware store to have the tank refilled.  The electric furnace uses power supplied by my local utility.  I am charged \$0.12 per kWh, and I estimated that the electric furnace costs about \$0.30 per hour to operate.

I continue to use my propane furnace to melt down and clean dirty scrap metal - metal that is oily or painted.  After pouring this cleaned metal into ingots it is suitable for use in the electric resistance furnace.  These two furnaces compliment each other.

\section{Summary}

The testing performed in this chapter confirmed that Furnace Version 2.0 operates very near its intended design power of 4000 watts and is capable of reaching useful operating temperatures within a reasonable time. 

Temperature measurements demonstrated that the insulating firebrick and ceramic fiber insulation system is highly effective, keeping most exterior surfaces only slightly above ambient temperature even when the furnace cavity is at 1000 \dg C. The highest exterior temperatures were observed at the lid-to-body interface, indicating that this area is the primary source of heat loss. 

Duty cycle measurements showed that the heating elements must operate for an increasing percentage of time as furnace temperature rises. 

Aluminum melting trials demonstrated that the furnace is fully capable of supporting hobby casting operations while maintaining precise temperature control. 

Energy consumption and operating costs were found to be low, making electric resistance heating an economical alternative to propane. Efficiency calculations indicated that only a portion of the supplied electrical energy reaches the aluminum charge, revealing opportunities for future improvement. 

Overall, the furnace met its design goals and proved to be a practical, economical, and effective tool for metal melting and heat-treating applications.

In the next chapter we will explore how these capabilities can be applied to practical metal casting projects.

Chapter Eight - Firing and Testing

Now that the furnace is assembled and completed it is time to break it in and begin testing it. 

A graphite saucer was placed on the furnace floor beneath the crucible. This prevents the crucible from bonding to refractory coatings, spilled metal, or debris on the furnace floor and allows the crucible to expand freely during heating.

Because this is a new crucible, it has to be tempered before use and the low temperature testing environment of this initial run is ideal for this purpose.  

 \section{Drying the Mortar and Adhesives}

For the initial run I set the PID controller setpoint to 100 \dg C and let it warm up and come to temperature.  I will let the furnace soak at this temperature for about two hours.  This will temper the new crucible and ensure that all refractories used in the construction are thoroughly dried.  The slow heat-up also reduces thermal shock to newly installed refractory materials.  Do not be surprised to see some smoke and some water vapor escaping from under the lid.  When you open the lid, you may encounter some scorch marks or staining from the adhesives curing.

After two hours I shut off the power and allowed the furnace to gradually cool down overnight.

The next morning I removed the tempered crucible and inspected all surfaces of the hot zone interior for cracks or any failed refractory.  There are some small cracks, but as long as they do not grow during the higher temperature runs to follow, I will be satisfied.

\section{Gradually Raising Temperature}

The following day I loaded the crucible and began the next phase of testing.  I entered 200 \dg C as the setpoint and let the furnace heat up.  When it reached 200 \dg C, I let it soak for an hour, and then I changed the setpoint to 400 \dg C.   I did not time these heat ups or estimate any kind of efficiency at this stage as I was only taking the furnace for a test drive.

At two hours I changed the setpoint to 600 \dg C.  I began to cautiously check the furnace shell temperature with my non-contact infrared thermometer and found that the surface temperature was just a little warmer than room temperature as I expected.  I also checked the control cabinet temperatures.

At the three hour point I raised the setpoint to 800 \dg C.  I was very interested in this temperature range because this is the upper limit of the temperature range I will frequently use for the casting of aluminum.   I opened the lid to inspect the interior, and as I expected, the contactor opened, and the elements de-energized.  The interior was glowing bright orange.

I closed the lid and depressed the Start Pushbutton again to resume the firing.

At four hours I changed the setpoint to 1000 \dg C.   Here the high temperature alarm began to flash and its buzzer began to sound.  I made the high temperature alarm setpoint 1000 \dg C because the heating element I used will melt and fail at 1149 \dg C.

\section{A Run at Full Temperature}

I took the alarm cutout switch to "cutout" to silence the alarm and I continued to hold the furnace temperature at 1000 \dg C for an hour.   After one hour I shut down the furnace and allowed it to cool to room temperature overnight.

\section{Tuning the Controller}

Because the Inkbird ITC-100VH controller may be installed in a furnace, a kiln, a heat treating oven, a smoker, or any other application that requires temperature control, It has a feature that allows it to determine the best Proportional, Integral, and Derivative (PID) values for your particular application \cite{inkbird_itc100vh}.

Before you begin, navigate through the controller menu to find the P, the I, and the D values.  Record these factory-set values.   If you later decide that the auto-tune values are poor, you can manually re-enter the factory values.

With the furnace at room temperature, change the controller setpoint to 750 \dg C.   Navigate through the controller menu until you come to CtrL.  Change the value to "AT" or 2.   The controller will display "AT" flashing when auto-tuning is active.

The controller will intentionally overshoot and undershoot as it measures heat up rate, cool down rate, thermal lag, and system response.   The tuning will take several heating and cooling cycles to complete the measurements.  Do not open the lid while the tuning process is active.  Do not interrupt the tuning process.  When the process is complete the "AT" will stop flashing.

When the auto-tuning process has completed, navigate through the controller menu to find the new P, I, and D values.  Record these values for reference and in case they have to be manually entered.

With these first testing runs completed and the controller tuned for the furnace it will be controlling, the furnace is ready for regular use.  The following chapter presents performance data collected during furnace operation, including heat-up rates, power consumption, temperature measurements, duty cycle observations, and melting trials.

Chapter Seven - Wiring It Together

The control system was designed using industrial control practices wherever practical. The use of DIN rail components, ferrules, branch circuit protection, grounding, and a dedicated safety chain improves maintainability, reliability, and operator safety.

This chapter describes the wiring of the 240 VAC power circuit, the 24 VAC control circuit, the alarm system, and the safety chain. It also explains the operation of the start and seal-in circuit used to control the contactor.

\section{Wiring the 240 VAC Power Circuit}

Power enters the control cabinet through the rear wall of the enclosure using 10 AWG four-conductor cable. Conductors L1 and L2 are connected to a two-pole 25 amp circuit breaker. The equipment grounding conductor is bonded to the control cabinet, while the neutral conductor is not used and is insulated with a wire nut inside the enclosure.

See Figure \ref{fig:furnacepower} to view a schematic representation of the power circuit.

The primary 240 VAC heater circuit is wired with 12 AWG THHN copper conductors rated for 90 \dg C service and 30 ampacity. Solid copper wire is used between the circuit breaker, contactor, and SSR. Flexible 10 AWG stranded high-temperature wire with fiberglass and mica insulation is used between the SSR and the heating elements. The stranded conductors are connected to the heating element terminals using stainless steel screws, washers, and nuts.

Ferrules are crimped onto the ends of stranded conductors wherever connections are made to the SSR, contactor, or lever-type wire connectors. The ferrules prevent strand damage, improve connection reliability, and reduce the likelihood of loose electrical connections.

The remaining 240 VAC branch circuits serving the control power transformer, PID controller, cooling fans, and volt-ohm meter are wired using 16 AWG stranded conductors. Since none of these loads exceed one ampere, this conductor size provides ample capacity. 

A combination fuse holder and distribution header is used to distribute power to the branch circuits and simplify wiring modifications and maintenance. 

The circuit breaker, contactor, SSR, and power distribution header are mounted on 35 mm DIN rail attached to a mounting plate in the rear of the control cabinet.

\begin{description}

    \item[25-Amp Double Breaker] A two-pole 25 amp 400 volt circuit breaker used for 

    circuit protection and isolation.  CNC Model YNB7-63N.

    \item[3-Pole Contactor with Auxiliary Contact] A three-pole normally open switch rated for 30 amps and 600 volts.  One normally open auxiliary contact is used for coil seal-in and indication. 24 VAC coil.  The contactor provides positive isolation from line voltage and is remotely closed by the control circuit.  Model CJX2-1810.

    \item[Solid-State Relay (on L2 only)] DIN rail mounted solid-state relay with heat sink.  3-32 VDC input.  24-480 VAC output.  Rated for 40 amps.  LCLCTC Model LCDS4048ZD3.  Only Conductor L2 is switched by the SSR. Positive isolation of both line conductors is provided by the contactor when the furnace is shut down.

    \item[Heater Elements] Kanthal alloy (FeCrAl) wire diameter: 1.2mm (0.047"); coil diameter: 7.2mm (0.28"); length: 800mm (31.45") each.  Rated voltage: AC220V, Rated Power: 3000W; Resistance: 15-16ohm. Withstands temperatures up to 2100°F.  Two are used in series in this project.

\end{description}

Although each heater element is sold as a 220 VAC, 3000 watt heater, the elements are operated in series in this design. The elements were shortened and adjusted to achieve the desired furnace resistance and power rating.

\section{Wiring the 24 VAC Control Circuit}

The control power circuit begins at the 240-VAC-to-24-VAC control power transformer. The transformer provides a reduced-voltage supply for the furnace control and safety systems. Using 24 VAC for control functions reduces the voltage present at switches and indicators that may be handled by the operator during normal operation and maintenance.

The 24 VAC supply is distributed to the safety chain, indicator lamp, and alarm circuit.

The 24 VAC control circuit is wired with 16 AWG stranded wire with ferrules crimped onto the ends for connection to terminals.  The ferrules prevent strand damage, improve connection reliability, and reduce the likelihood of loose electrical connections.

See Figure \ref{fig:furnacesafety} to view a schematic representation of this circuit.

\subsection{24 VAC Safety Chain}

The safety chain is a series circuit consisting of five normally closed safety switches and one normally open momentary start pushbutton used to energize the contactor coil. When the contactor coil is energized, the contactor closes and permits power to reach the heating elements. If any normally closed safety device opens, current flow to the contactor coil is interrupted and spring pressure causes the contactor to open, removing power from the heating elements.

The safety devices included in the chain are:

\begin{description}

    \item[Emergency Stop Button] This is a large red normally closed momentary push button mounted on the top of the control cabinet.  When the operator presses this button the safety chain opens and de-energizes the contactor coil causing the contactor to open and isolate the electric heating elements.

    \item[Cabinet Over-Temperature Switch] A small normally closed bimetallic snap-action thermal switch mounted to the control cabinet wall set to automatically open at 90 \dg C.  When the switch opens the safety chain opens and the furnace shuts down.  A 90 \dg C switch was selected to protect wire insulation and components in the cabinet.

    \item[Lid Limit Switch] A normally closed switch that provides a lid safety interlock. Opening the furnace lid causes the switch to open, interrupting the safety chain and de-energizing the contactor. This removes power from the heating elements before the operator can access the furnace interior. The lid limit switch is installed inside the furnace body shell.

    \item[Dump or Anti-Tilt Switches] Two normally closed tilt switches are wired in series, and will open whenever their tilt is more than approximately 15\dg \ from vertical.  This causes the furnace to shut down if the furnace is accidentally tipped or overturned.  The dump or anti-tilt switches are mounted inside the furnace body at locations 90\dg \ from each other.

    \item[Start Push Button] A normally open momentary push button mounted on the front panel used to remotely energize the contactor coil and initially start the furnace.

    \item[Indicator Lamp] A panel-mounted green indicator lamp is connected in parallel with the contactor coil. The lamp illuminates whenever the contactor coil is energized, providing the operator with a visual indication that the safety chain is intact and the heating elements are capable of being energized.

\end{description}

\subsection{24 VAC Alarm Circuit}

A combination flashing LED indicator and audible alarm module is mounted in the top of the control cabinet. The alarm is controlled by a relay internal to the PID controller that closes when a programmed alarm setpoint is reached.

See Figure \ref{fig:PID-SSR} and Figure \ref{fig:furnacesafety} for schematic representations of the alarm circuit.

Control power is routed through a cutout toggle switch and connected to Terminal 12 of the PID controller. The alarm device is connected to Terminal 11, which provides a normally open relay contact. When the alarm relay closes, power is supplied to the alarm module.

A momentary pushbutton is wired in parallel with the PID alarm relay contact to permit alarm testing before furnace start-up. This feature allows the operator to verify proper alarm operation without requiring the furnace to reach the alarm setpoint.

\section{Starting and Seal-In}

The starting sequence will be explained now because it may not be apparent to readers how a momentary push button can start the furnace and permit continuous operation.

When all switches in the safety chain are closed (or reset) the furnace is able to be started.  The operator presses the normally open momentary switch on the front panel.  This completes the 24 VAC circuit to the contactor coil.  When the contactor coil energizes it closes the contactor.  The three main contacts close and the normally open auxiliary contact closes.  24 VAC is connected from the upstream side of the Start button to Terminal 13 of the aux contact.  Terminal 14 of the aux contact is connected to the contactor coil.  

While the start button is pressed, 24 VAC has two paths to the contactor coil -- through the Start button, and through the auxiliary contact.  When the Start button is released the aux contact continues to provide a current path to the contactor coil.  The contactor coil will remain energized and hold the contactor closed until a switch in the safety chain opens and removes the supply of current to the coil.

Following a shutdown event, the furnace will remain de-energized until all safety chain devices have returned to their normal condition and the Start Pushbutton is pressed again. This arrangement helps protect the operator from exposure to energized heating elements and prevents the furnace from automatically restarting after a power interruption.

Chapter Six - Making the Shell and Lid

 \section{Overview}

The furnace body and lid provide the structural framework that supports the heating chamber, insulation, control cabinet, sensors, and lid lifting mechanism. In addition to containing the thermal components of the furnace, the body must withstand the mechanical loads imposed during operation and maintenance. This chapter describes the fabrication of the steel shell, installation of insulation, construction of the lid, and integration of the supporting mechanical systems.

\section{Cutting the Drum}

The drum could not be cut until the heating chamber assembly was complete because the final height of the floor, insulation board, and firebrick structure determined the required shell height. \index{furnace shell!height}

The required shell height was measured and marked around the circumference of the drum using a silver paint marker.  I drilled a 1-inch diameter hole with the hole saw tangent to this line.  Then I easily cut around the drum with my Wen 3650 4.0 amp swivel head variable speed electric metal shear.   \index{steel drum!cutting}

Next I had to decide how thick the lid would be and measure down from the top.   At this height I marked out the cutting path again and cut around with the shears.

The shell supports the floor of the furnace and keeps the ceramic insulating blanket confined around the heating chamber.

\begin{figure}[htbp]

    \centering

    \includegraphics[width=0.8\textwidth]{images/body-lid}

    \caption{The steel drum cut into sections for the body and the lid.}

    \label{fig:body-lid}

\end{figure}

\section{Reinforcing the Shell and Lid}

The shell needed to be reinforced to handle some of the mechanical loads that it is expected to carry.  I used cold rolled steel in two-inch and three-inch widths to strengthen it.  I built a column along the back side of the shell to carry the lid lifting mechanism.  A three-inch wide strip of 1/4-inch CRS was fastened to the shell with 1/4-20 UNC bolts and rivnuts.  I welded a bar of 1/4-inch thick CRS to the center of this flat strip to make a stiffening rib.  This rib tapered from 3-inches wide at the base to 1-inch wide at the top.  \index{furnace shell!reinforcement}

The bottom was reinforced where the lifting mechanism pivot had to be fastened to the drum.  This was simply a 1/4-inch-thick bar attached to the bottom with rivnuts like the back column.

A backing plate was installed on the front of the drum to reinforce the area where the control cabinet would be supported.

Finally, the lid needed reinforcement around the perimeter, at the rear where it would be connected to the lifting column, and two backing plates where the turnbuckle pad eyes were to be bolted on.

\section{Penetrations for Power and Sensors} 

Two holes were cut with a hole saw to allow power and sensor wiring to pass between the control cabinet and the furnace shell. The heater power conductors were routed through one opening, while the thermocouple wiring and connections associated with the lid interlock and anti-tilt safety switches were routed through the second opening. Strain-relief fittings were installed in both openings and connected to flexible conduit.

Two paths were created so that low voltage and sensing lines could be separated from the power circuit to reduce the possibility of interference. \index{furnace shell!penetrations}

\section{Insulating the Lid}

The lid has four insulating firebrick set on their ends to hold the spacing from the top of the drum to the insulating fiberboard.  The space between the insulating fiberboard and the top of the shell is filled with ceramic fiber blanket insulation.  The fiberboard is held in place with metal clips screwed into the side of the steel shell.

This construction is a change from Furnace Version 1.0.  The legacy furnace had a layer of firebrick in the lid, and while this was not very heavy, the new construction method reduces lid weight by 8 pounds. \index{furnace lid}

\section{Pedal-Actuated Lid Lift}

A foot-operated lid lifting mechanism was incorporated into the design to improve both furnace efficiency and operator convenience. During melting operations, the lid must be opened periodically to inspect the charge, add additional metal, skim dross, or insert fluxes. With Furnace Version 1.0, the lid had to be lifted completely off the furnace and placed on a nearby firebrick support before these operations could be performed. Afterward, the lid had to be lifted a second time and returned to the furnace.

The new lifting mechanism allows the lid to be raised and swung aside in a single motion. This reduces the time that the heating chamber remains open, minimizing heat loss and shortening recovery time after the lid is closed. The mechanism also reduces the amount of lifting required by the operator and improves overall ergonomics during furnace operation.

A foot-operated mechanism allows the operator to keep both hands available for repositioning the lid and handling tongs and crucibles. \index{lid lift!foot-operated}

The lid lifting column and pivot mechanism were intentionally located on the side of the furnace opposite the control cabinet. As a result, the lid always swings away from the controls when opened. This arrangement prevents the hot face of the lid from being positioned above the control cabinet, reducing the likelihood of exposing switches, wiring, indicators, and electronic components to unnecessary radiant heat. The configuration also provides the operator with unobstructed access to the controls while the furnace is open.

Below the control cabinet at the bottom of the furnace shell the pedal was installed.  The pedal is welded to the top of a bar with an I cross-section.  The top and bottom flanges are ground away where the bar passes through the pivot point.  Here a brass bushing was pressed into the web and a smooth shoulder bolt passing through the pivot flanges makes a fulcrum for the lever.  At the back of the furnace the bar lifts a solid one-inch diameter bar of round stock to lift the lid.  

\section{Thermocouple Installation}

A small hole was cut into the top of the drum and an electric junction box was screwed into the inside of the lid.  A cover was screwed over the box for access.  This box was used to mount the thermocouple.  The thermocouple is covered with a protective ceramic sheath and it protrudes from the face of the fiberboard into the heating chamber.  The sheath is fixed to the fiberboard with sodium silicate adhesive. \index{thermocouple!installation}

Because the thermocouple hangs below the lower lid surface, the lid lifting mechanism must lift the lid high enough for the thermocouple to clear the sides of the heating chamber.  This is accomplished by a pin installed in the lift bar that follows an L-shaped slot machined into the surrounding guide pipe.  Following this L-shaped path, the lid cannot be rotated until the lid has been lifted to the top of the slot where the slot turns 90 degrees and runs horizontally around 90 degrees of the pipe.  At the end of the slot a small detent is filed to hold the lid in position when it is open.

\subsection{Lid Interlock and Safety} 

A limit switch is mounted near the top of the furnace shell.  The limit switch is an important safety device that interrupts power to the contactor coil if the lid is opened.  When the contactor coil is de-energized, the contactor opens and removes power from the heating elements.  The limit switch has a plunger that protrudes above the insulating fiber board.  This spring-loaded plunger is depressed by a metal tab on the lid when the lid is in position and lowered in the shut position.  \index{lid interlock} \index{limit switch} \index{safety chain}

\section{Completed Furnace Body and Lid}

After fabrication was complete, the sharp edges created when the drum was cut were covered with metal-reinforced vinyl edge trim. This trim reduces the risk of cuts and abrasions during operation and maintenance while also providing a more finished appearance. The edge trim was installed around the upper edge of the furnace body and the lower edge of the lid wherever exposed sheet metal edges remained accessible to the operator.

With the shell reinforced, the insulation installed, the thermocouple mounted, and the lid lifting mechanism adjusted, the furnace body and lid assembly were complete. The resulting structure provides mechanical support for the heating chamber while minimizing heat loss and facilitating safe operation. The following chapter describes final assembly of the furnace and integration of the body, heating chamber, and control system.

Chapter Five - Heating Chamber Construction

This chapter describes the construction of the furnace heating chamber, including the insulating firebrick hot face, heating element grooves, furnace floor, ceramic fiber insulation, and initial testing before final installation.

\section{Overview of the Heating Chamber}

Dave Gingery made the heating chamber of Lil Bertha from castable refractory cement and he molded the element groove into the sides of the chamber with rubber hose.  The refractory was held in place with a temporary internal form and the metal sides formed by galvanized metal.  \index{Gingery, David} \index{Gingery, David!Lil Bertha} \index{refractory, castable}

He made the furnace in three parts - the base standing on short pipe legs, the body, and the lid.  He attached his range control to the side of the body in a small box.  There were advantages to this set up:  you could stack multiple bodies for a taller heating chamber, you could remove the base and use the furnace for melting out wax.

Dan Hartman quickly learned how difficult it was to form the groove, to remove the temporary form from the interior, and how unreliable a range control would be for temperature control.  \index{Hartman, Dan}

Dan improved on Gingery's design by using commercially available HVAC duct for the shell of the base, the body, and the lid.  Dan overcame the problem of the molded element groove crumbling and being difficult to form by cutting grooves into insulating firebrick arranged in a hexagon around the heating chamber.  He made his own homemade refractory material out of furnace cement mixed with perlite.   Dan designed and built his own electronic controller for heater control.  

On Version 1.0 of my shop furnace, I incorporated Dan's idea of using firebrick for the hot face, but I also realized that I needed to keep the furnace mass low to reduce heat transmission.  I used a commercially available steel drum for the body shell and I placed it on a drum dolly to make it portable in the shop. \index{steel drum} \index{drum dolly}

The heating chamber has a groove cut into the side and the heating element coil is installed in the groove.  In the Gingery design the heating element spirals around the heating chamber twice.  In Dan's design, the element circles around the heating chamber six times, but it does not follow a spiral:  he cut ramps in the terminal brick that guides the element from level to the next.  At the top and bottom of the terminal brick there are holes drilled for the heating element leads to pass through. \index{terminal brick}

On Version 1.0 of my shop furnace, the heating element groove follows a continuous incline on the faces of the eight firebricks.  There is a terminal brick and it has holes at the top, bottom, and midlevel.  The top and bottom holes connect the heater to the power supply.  The middle hole is used to connect the two coil segments to each other in series.

\section{Materials Used}

I do not use castable refractory at all in my furnace design.  I use Vulcan refractory furnace cement to bond the firebricks together and seal the joints.   The vertical firebricks are bonded to each other at the edges and they are clamped together with encircling worm drive band clamps.   \index{refactory cement} \index{band clamps}

The vertical heating chamber walls are not bonded to the floor.  The vertical part of the chamber floats on the floor so that it can be disassembled for element replacement.  The chamber is kept centered by bolts screwed in and supported by the furnace shell.

The face of the firebricks are painted with ITC-100 HT to protect the firebrick.  This coating is rated up to 5000 \dg F.  This coating reflects heat back into the furnace which improves heat up times and raises furnace efficiency. \index{refractory coating} \index{ITC-100 HT}

I used insulating ceramic fiberboard on the lid and below the firebrick floor between the firebrick and the bottom of the steel drum. \index{fiberboard, insulating}

The space behind the vertical firebricks is loosely filled with ceramic fiber insulation blanket. \index{ceramic fiber blanket}

\subsection{Insulating Firebrick}

The 2600 \dg C insulating K26 firebrick (IFB) that I used are made by Morgan Advanced Materials.  I purchase these locally at Joe Moore Company, a boiler and furnace materials supplier, to save the cost of shipping.  \index{firebrick} \index{Joe Moore Company} \index{Morgan Advanced Materials}

The IFB are soft and can be easily cut with wood working tools.  Care must be taken because the corners can chip easily and if you are not careful the brick can crack and break.  If the break is clean, the two halves can be cemented back together with refractory cement.  \index{firebrick!broken}

The IFB have the following dimensions:  9 inches tall, 4.5 inches wide, and 3 inches thick. \index{firebrick!dimensions}

\subsection{Ceramic Fiber Board}

The ceramic fiberboard that I use below the firebrick floor is rated up to 2300 \dg F.  It has a density of about 20 pounds per cubic foot, it is one-inch-thick, and it is easily cut with a knife.  It has low thermal conductivity and is almost impervious to thermal shock.  \index{fiberboard, insulating}

\subsection{Ceramic Fiber Insulation}

The ceramic fiber insulation blanket is rated up to 2400 \dg F.  This material is easily cut with a knife.  It has low thermal conductivity making it a great insulator.  \index{ceramic fiber blanket}

\textbf{Safety Note: } This material is asbestos-free, but the fibers can be hazardous when they are airborne, so you must wear gloves and an N100 particulate mask when working with this material in a well-ventilated space. 

Use the vacuum cleaner - not the broom - to clean up loose fiber after cutting.  Wear long sleeves -- this material will irritate your skin just like fiberglass insulation.   \index{ceramic fiber blanket!hazards} \index{cleaning up}

This material is more hazardous after it has been exposed to high temperatures because the fibers change their geometry and become needle-like lodging deeply in the lungs.  When disassembling the furnace for maintenance it is crucial that the worker use respiratory protection.

\begin{figure}[htbp]

    \centering

    \includegraphics[width=0.8\textwidth]{images/firebrick-arch}

    \caption{Firebrick geometry used to form the heating chamber.}

    \label{fig:firebrickarch}

\end{figure}

\begin{figure}[htbp]

    \centering

    \includegraphics[width=0.8\textwidth]{images/element-grooves}

    \caption{Heating element groove as it transitions from brick to brick.}

    \label{fig:element-grooves}

\end{figure}

\section{Preparing the Firebrick}

Dan Hartman cut both inside corners off of his firebricks to make an arch around his heating chamber.  I was playing around with the geometry and I discovered that only one corner needs to be cut off of each firebrick to create the brick arch.  That is what I did.  I cut off the corners at a 45 \dg angle using my table saw starting about 0.75 inch from one side.  See Figure \ref{fig:firebrickarch}. \index{firebrick!cutting} \index{firebrick!arch}

I created a cardboard template for the heating element grooves and traced the groove path onto the face of the bricks.  

\subsection{Cutting Firebrick}  

The insulating firebrick are very soft and easy to cut.  They can be cut with a wood saw, but the bricks are very abrasive and they will quickly dull your saw.  I purchased an inexpensive hand saw that will be sacrificed for this job. \index{firebrick!cutting} \index{saw, sacrificial}

\subsection{Routing Heating Element Grooves}

After drawing the heating element groove outline onto the face of the firebricks with pencil, I used a router bit in a small trim router running along a wooden fence clamped over the brick to cut the groove.   The groove is about 3/8 inches wide and 3/8 inches deep.

The firebrick are very abrasive and they will ruin your router bit.  I used an inexpensive router bit that was sacrificed for cutting the heater groove.  After the groove was cut I smoothed the edges and transition from brick to brick with a round file. See Figure \ref{fig:element-grooves} \index{router bit, sacrificial}

\section{Heating Element Design}

The heating element is formed from coiled resistance wire and it follows a spiral groove around the heating chamber so that heat is radiated from the chamber walls uniformly.  I am using two elements joined in series. \index{heating element!design}

\subsection{Required Resistance}

For this project I require a heater with about 14.4 ohms electrical resistance (See Equation \ref{eq:current1}).  Each element should have about 7.2 ohms of resistance.   I gently stretched the element out so that the coils were not touching each other and checked the resistance again with my Fluke portable volt-ohm meter.  \index{resistance, electrical}

The element has a length of uncoiled wire on each end.  This is not included in my resistance measurement -- only the coiled section is measured for resistance.

The element has a resistance of 9 ohms before it is shortened.  The resistance of the element is proportional to the length of the resistance wire. Since the coil is manufactured from wire of uniform diameter and composition, the required element length can be estimated using the ratio of resistance to wire length: \index{electrical resistance!ohms per inch}

\begin{equation}

    \frac{R_1}{R_2} = \frac{L_1}{L_2}

\label{eq:RLratio}

\end{equation}

\begin{equation}

    \frac{9\ \Omega}{7\ \Omega} = \frac{15\ inches}{L_2\ inches}

\label{eq:newlength}

\end{equation}

Solving for $L_2$ yields:

\begin{equation}

L_2 = 11.7\ \text{inches}

\label{eq:L2}

\end{equation}

The new length is about 3.3 inches shorter.  I cut the wire off one inch at a time checking the resistance with my volt-ohm meter after each cut.  \index{heating element!cutting}

The element wire has about 3 inches of uncoiled wire on each end.  I bend this over to double it on itself (and reduce its resistance) and twist this together tightly.  This will be the wire segment that passes through the hole in the terminal block and is connected to the electrical supply.

\begin{figure}[htbp]

    \centering

    \includegraphics[width=0.8\textwidth]{images/initial-test}

    \caption{The first reduced power test of the heating assembly.}

    \label{fig:initial-test}

\end{figure}

\subsection{Stretching the Elements}

I measured the heating element groove length by pressing a cotton rope into the groove, marking the end with a black marker.  I pulled the rope out and measured the length.  I found it to be 130 inches long.  I have two elements to stretch out and fill the entire 130 inches of groove.  The elements have been shortened to 11.7 inches. \index{heating element!stretching}

Each element has to be stretched to 65 inches long.  

I put the uncoiled tail into my bench vise and I put the end of my tape measure into the vise jaw next to it.  I slowly back up while pulling and stretching out the coil as the tape measure extends out.  Care must be taken to stretch the element uniformly so that the spacing between adjacent turns remains reasonably consistent. Doing this will prevent hot spots from being created.  I stop when I have stretched the element to 65 inches long.  I repeat this process for the second element. \index{heating element!hot spots}

\section{Heating Element Installation}

I push the doubled up wire end of the first element through the lower terminal brick hole and then I begin to carefully push the coil into the groove and work my way up to the halfway point (about 3 times around the heating chamber).  At the halfway point, the other doubled up wire end is pushed through the middle hole in the terminal brick. \index{heating element!installation}

I push the doubled up wire end of the second element through the middle hole in the terminal brick and then carefully push the coil into the groove and work my way up to the top.  At the top the doubled up wire end is pushed through the top hole.

In the middle of the terminal brick where the two element wire ends are pushed through the brick, the two elements are joined with a stainless steel bolt, washers, and nut.

Where the element crosses the firebrick corners it tends to lift up out of the groove -- the groove sections are linear, but the element tends to form a circular shape.  To ensure that the element remains in the groove, I pin the element in place with U-shaped pins of Kanthal wire. \index{heating element!pinning}

\section{Heating Chamber Testing}

Before final installation into the furnace body, the completed heating chamber assembly was subjected to an initial low-power test. The assembly was placed on two insulating firebricks to elevate it above the work surface and minimize heat transfer. \index{testing!low power}

The heating element was temporarily connected to a 120 VAC power source. Since the heating elements were designed for operation at 240 VAC, this reduced-voltage test limited the power output and allowed the assembly to be heated gradually. The purpose of the test was to remove any remaining moisture from the refractory materials, verify the integrity of the heating elements and electrical connections, and identify any obvious construction defects before the furnace was fully assembled. See Figure \ref{fig:initial-test}. \index{testing!purpose}

During the test, the heating element reached a dull red color and was energized for approximately one hour. No damage to the heating elements, refractory cement, or firebrick was observed. This preliminary cure cycle provided additional drying of the refractory materials before the furnace was placed into service at full operating power.  

\textbf{NOTE:} After the initial firing, the Kanthal heating element will have been exposed to elevated temperatures and will no longer possess the same ductility it exhibited during installation. The element becomes progressively more brittle with use and may fracture if bent sharply or handled roughly. If maintenance or repairs require the heating element to be moved, it should be handled carefully to avoid accidental breakage. \index{Kanthal} \index{heating element!used}

\section{Heating Chamber Floor Construction}

The furnace floor was constructed by cementing several insulating firebricks together along their edges to form a single assembly. After the refractory cement had cured, the assembly was cut into a circular shape using a saw. A coat of ITC-100 HT was then applied to the surface exposed to the heating chamber. \index{firebrick} \index{ITC-100 HT} \index{heating chamber!floor}

Two layers of one-inch-thick ceramic fiberboard were cut into circles and placed in the bottom of the steel drum. The completed floor assembly was lowered into the drum using lifting straps. After the floor was positioned, the straps were carefully withdrawn from beneath the brick and removed from the furnace shell. \index{fiberboard, insulating}

\begin{figure}[htbp]

    \centering

    \includegraphics[width=0.8\textwidth]{images/completed-HC}

    \caption{The heating chamber and insulation blankets before being closed.}

    \label{fig:completedHC}

\end{figure}

\section{Completed Heating Chamber}

The vertical firebrick assembly was carefully lowered into the drum and arranged on the floor.  Carriage bolts were screwed in from the side walls so that their ends were in contact with four of the bricks.  These bolts would keep the heating chamber centered in the shell. \index{heating chamber!assembly}

The heating elements were connected to high temperature hook-up wire with stainless steel bolts, washers, and nuts.  The high temperature hook-up wire passes through the furnace shell and is routed to the control cabinet where it is connected to the SSR and the contactor. \index{high-temp hook-up wire}

The lid safety interlock switch was mounted to the steel furnace shell and connected to the safety chain in the control cabinet using 16 AWG wire. The switch is wired using its normally open (NO) contacts and closes only when the lid is fully closed. \index{lid interlock}

Two anti-tilt switches were also mounted to the furnace shell. These switches are wired in series with each other and with the lid safety interlock switch. Any switch opening will interrupt the safety chain and de-energize the contactor, removing power from the heating elements. \index{anti-tilt switches} \index{safety chain}

To support future performance testing and data collection, four thermocouple junctions were cemented to the rear faces of selected firebricks.  Their wires terminate at connectors mounted on the side of the furnace shell, allowing a portable thermocouple reader to be connected when rear-face brick temperatures are to be measured. \index{thermocouple!static devices}

After these connections were completed, ceramic fiber insulating blanket was cut and carefully inserted into the annular space between the heating chamber and the steel drum.  The insulation was installed loosely enough to avoid excessive conductive heat transfer, yet densely enough to prevent significant air circulation and convective heat loss.  See Figure \ref{fig:completedHC}. \index{ceramic fiber blanket!installation}

Finally, a wide ring of ceramic fiberboard was cut and fitted to the top of the body to seal off the annular internal space.  This ring seals the annular insulation space, helps keep the ceramic fiber blanket clean, and reduces the release of loose fibers during operation and maintenance. \index{fiberboard, insulating}

\subsection{Initial Electrical Checks}

Before connecting the heating element assembly to the 240 VAC power supply, several electrical checks were performed. These checks were intended to verify the integrity of the heating elements and confirm that no unintended electrical paths existed between the heater and the furnace shell. \index{electrical testing}

Using a Fluke portable volt-ohm meter, the resistance of the assembled heater was measured at the terminal connections. A resistance of 14.9 ohms was obtained, compared to the design value of 14.4 ohms. The small difference is likely attributable to the additional conductor length introduced by the uncoiled element leads and measurement uncertainty. \index{resistance, electrical!measurement} \index{heating element!resistance}

This value was considered acceptable (about a 3.5\% difference). A reading of OL (open line) would have indicated a broken heating element or an open connection. Conversely, a reading near zero ohms would have indicated a short circuit within the heater assembly.  \index{resistance, electrical!continuity}

Next, the resistance between one heater terminal and the grounded steel furnace shell was measured. The meter indicated OL, corresponding to infinite resistance. This confirmed that no electrical connection existed between the heating element and the furnace body.  \index{resistance, electrical!ground}

Having verified that the heater assembly was electrically sound, the heating elements were connected to the SSR and contactor in preparation for the initial low-power test.

With the heating chamber complete and initial testing successfully concluded, the project was ready to proceed to final assembly. The following chapters describe the integration of the heating chamber with the furnace body, lid, control system, and supporting hardware. Detailed performance testing, operating data, and efficiency measurements will be presented later in this document.

Chapter Four - Building the Electric Controls

The control system was designed around three primary objectives: safety, reliability, and ease of maintenance. Safety is achieved through the use of low-voltage control circuits, electrical interlocks, and positive contactor isolation. Reliability is enhanced through conservative component selection and thermal management. Ease of maintenance is provided through DIN rail construction, terminal blocks, and clearly defined wiring paths.

The heater is powered by 240 VAC single-phase electricity supplied by a 30 amp branch circuit in my garage.  As you read in Chapter Two, Equation 2.3, the heater draws 16.7 amps.  To safely carry this load while complying with common wiring practices for a 30-amp branch circuit, 10 AWG copper conductors are used.  The furnace is connected to the branch circuit by a 25-foot-long 10/4 AWG cable terminated with a NEMA 14-30 plug.  

One of the conductors is connected to ground.   One of the conductors is connected to neutral. Two conductors, designated L1 and L2, are connected to opposite phases of the residential split-phase electrical service. Approximately 120 VAC exists between either line conductor and neutral, while approximately 240 VAC exists between L1 and L2.  The neutral conductor is not used by this design and is terminated with an insulated cap inside the control cabinet.

Control Cabinet

To be sure that all of the components needed in my control scheme would fit into the control cabinet, I built a model controller with the components mounted on DIN rails on a piece of scrap wood.   Satisfied that I could assemble the controller without components interfering with each other, I ordered a NEMA stainless steel electrical enclosure that measures 12 inches by 12 inches by 8 inches deep with a locking door on the front of it.  

Aluminum handles were mounted to the sides of the cabinet so that the user may tilt and reposition the cabinet for optimum viewing.

Penetrations were cut into the door and cabinet to mount the PID controller, the volt-amp meter, the Start button, the heater energized lamp, the E-stop button, the alarm module, the alarm cutout button, the alarm test button, the cooling fans, and the fuse holders.

The DIN rail is mounted horizontally on a steel plate that is installed on stand-offs in the back of the cabinet.   The lower rail holds the 240 VAC power components -- the 25 amp circuit breaker, the contactor, a terminal block for 240 volt loads, and the SSR.  The upper rail holds terminal blocks for the 24 VAC control and safety circuit.

The tilt and rotate feature is facilitated by using a re-purposed PanaVise mounted to the furnace shell, and a 1-inch steel pipe and floor flange mounted to the back of the steel cabinet.  240-volt high temperature hook up wire going to and from the heater pass through a hole in the back of the cabinet and through the center of the Panavise joint.  Shielded thermocouple wire passes through this neck joint, as well as 16 AWG wire for the lid limit switch and the dump switches.

Figure \ref{fig:furnacepower} illustrates the distribution of 240 VAC power within the control cabinet.  L1 and L2 enter the cabinet from the rear and they are connected to CB1, the 25-amp circuit breaker.  The ground wire is firmly bonded to the side of the cabinet.  From CB1 12 AWG solid wire is used.  It goes to a two-in, four-out lever-type connector.

One side of this fork goes to L1 and L2 terminals on Contactor K1, and the other branch goes to a two-in, ten-out lever-type connector.  The remaining 240 VAC loads are distributed from that connector.  

Power leaves contactor K1 through terminals T1 and T2. The conductor connected to terminal T1 is routed to terminal L1 of the SSR. The conductor connected to terminal T2 passes through the current transformer before continuing to the heating element.

Terminal T1 on the SSR is connected to the heater with another length of 12 AWG high-temperature hook-up wire.

PID Controller

I selected the Inkbird Model ITC-100VH PID controller for this project.  This controller has operated satisfactorily on Furnace Version 1.0 and I had a spare to use on this project.  

This controller is mounted on the front panel of the control enclosure.  It displays the current setpoint temperature value - SV (setpoint temperature) and the current process temperature value - PV (actual temperature).   It has a small LED that is lit when the SSR is closed.

240 VAC is connected to the controller on Terminal 9 and 10. The controller is powered directly from the 240 VAC supply and internally converts this voltage to the low-voltage signals required for measurement and control \cite{inkbird_itc100vh}.

The Type K thermocouple is connected to Terminal 3 and 4, and Terminals 4 and 5 are tied together.

The controller supports the use of an alarm module via an internal relay connected to Terminals 1, 11, and 12.   24 VAC is connected to Terminal 12, and the alarm module is connected to Terminal 11.  See Figure \ref{PID-SSR} and \ref{fig:furnacesafety}.

Solid-State Relay

A DIN rail-mounted 40-amp SSR with integrated finned heat sink was selected because of its slim profile and space savings.  The SSR is switched by the PID controller using a 12 VDC signal conducted to the positive (+) and negative (-) terminals on the control side of the SSR.   

Although the heater load is approximately 16.7 amps, a 40-amp SSR was selected to provide additional thermal margin and improve reliability.  See Figure \ref{PID-SSR}.

Using a Contactor for Safety

A big reason for this controller redesign was the discovery that the SSR has significant current leakage even when it is off.  This presents an electric shock hazard to the furnace operator.  

An electric contactor opens the circuit with an isolating air gap.  The contactor is remotely closed with 24 VAC control power via a momentary push button switch and maintained closed with a seal-in circuit using normally open contacts 13 and 14.

A safety chain of normally closed switches conducts 24 VAC to the contactor coil.  These switches include the E-Stop button, a cabinet over-temperature switch, a pair of anti-tilt switches, and a lid limit switch. When any of the switches open, the supply of control voltage is interrupted and the contactor opens under spring force.  After the cause of the trip has been corrected the contactor is closed again by pressing the Start momentary switch. See Figure \ref{fig:furnacesafety}.

Alarms and Indicators

The alarm module consists of a buzzer and flashing LED lamp.  It is mounted in the top of the cabinet.  The alarm is powered by the 24 VAC control circuit.  The alarm is actuated by a relay in the PID control that is closed when the pre-set alarm setpoint is reached.  A panel-mounted switch allows the alarm to be temporarily disabled without changing the controller settings. The alarm can be tested for function with a momentary switch push button that by-passes the controller and cutout switch.  The alarm should be tested before each furnace run to ensure it functions, and to be certain that the cutout switch has been returned to the enable position.

Indication that the contactor is closed is given on the front panel by a lamp wired in parallel across the contactor coil.  When the coil is energized the lamp is illuminated.   When the contactor trips, the lamp is extinguished.

In addition to the indications provided by the PID controller, a small electronic meter displays instantaneous voltage and current on the front panel. See Figure \ref{fig:furnacesafety}

The control system for Furnace Version 2.0 was designed with safety, maintainability, and reliability as primary objectives. The use of low-voltage control circuits, safety interlocks, positive contactor isolation, alarm annunciation, and organized DIN-rail construction represents a significant improvement over Furnace Version 1.0. Together, these features help reduce the likelihood of equipment damage and operator injury while simplifying troubleshooting and future modifications.