Tin Can Boiler: Materials Experiment (Hard Science)

(click to see larger image of rough boiler plans)

BIG HUGE DISCLAIMER! – Let me start off this entire article by saying there is a very real danger from working with boilers and you should not attempt this project unless you know exactly what you are doing and accept all risk and responsibility for your actions. I’m not joking about this: this is hard science, kids, and this blog is for documentation of my results. I am obsessive-compulsively careful in regards to these things, and have taken great care to research each step of the way. Do not use my notes as a shortcut. You could get killed by an explosion doing this stuff. You have been warned. Repeatedly. I assume absolutely no risk or responsibility if you choose to try to do this despite all my warnings. That said…

Yes. I’m going to make a steam engine. This is hard-science, and I admit, it feels pretty good. After repeatedly researching a number of engineering forums and sites, I sketched out the above plans in my moleskine. My first steam engine is going to be a very simple affair only for the purpose of turning a wheel. The engine test will happen later, and will first be tested with compressed air. The general idea is to use a 30-psi or less boiler. The simple engine I wish to build requires less than 5-psi, which means I should be able to get a substantial amount of work from my boiler if successful.

The boiler itself calls for a brick base and surrounding structure to create the oven, with a semi-open front and a ventilation pipe allowing for exhaust fumes and airflow. The boiler itself will be made of a tin-can. The bottoms of two aluminum cans will serve as the coal-cradles. Two pipes will each lead to rubber hoses: the first to a pressure release valve which will kick in automatically at about 30psi (or less), the second will head to the engine, first passing a coupling. One path of the coupling will be a low-gauge pressure valve so that the pressure can be monitored. The second will lead to a valve that can be opened to release the steam into the engine itself. A sliding tray will allow for the coals to be quickly removed from the oven if need be, and a water refill cap will likely be installed at the top in later models, though this first one will largely be so that I can learn and make mistakes along the way.

That said, there is a great deal of testing to be done if I expect the engine to work, and to not get killed while doing it. The very first of these tests is even to see if my idea to use quick-lighting “hookah charcoal” will indeed boil water, if tinplate steel will allow for enough transfer of heat from the coal to the water, and if any pressure can be built up. This, of course, requires an experiment for science!


  • Will water in a tin can boil from the heat of quick-lighting charcoal tablets?
  • Will enough pressure be generated to push a can upward?
  • How much water will be boiled off or evaporate?


  • It will require two charcoal tablets for the water to boil.
  • Enough pressure will be generated, provided the seal is tight enough and the friction is low enough.
  • Because of the low heat and closed system, I expect very little water to boil off.

(materials used, minus the meat thermometer and a more precise measuring cup.)

Materials List:

  • 1 tinplate-steel food tin, with the label peeled off.
  • two aluminum cans.
  • 1 set leather work gloves (fingers and palm at least).
  • 5-sided metal box.
  • needle-nosed pliers.
  • scissors.
  • accurate measuring cup(s) with 1/2 cup water.
  • quick-light charcoal tablets (we used “Sultan Charcoal” because we were assured it burned hottest of the available selection).
  • lighter.
  • paper-towle.
  • meat-thermometer.
  • a well-ventilated, safe work area with shield that can be put into place.

(steps 1-3)

Steps to Reproduce: (note: skipped step numbers are cosmetic only, to post photos throughout the experiment)

  1. Bend the two long-walls of the metal box towards one another until the tin can can rest atop it without being in danger of falling over, while still leaving room for the aluminum can bottom.
  2. Cut the bottom off of the aluminum can, leaving just enough room so that the combined height of the aluminum can bottom, plus the charcoal, rests just under the tin can when placed in the box.
  3. Dull the edges of the cut tin can bottom to help prevent getting cut and provide a more even surface. At first I thought I’d use the pliers but then found scraping it along the concrete patio worked much faster.
  4. (steps 5-6)

  5. Pour the 1/2 cup water into the tin can.
  6. Place the cradle in the box, and the charcoal into the cradle (for the first test, only one tablet will be used).
  7. Place metal box into a prescribed “safe work area.” For the purposes of this experiment, A closeable stainless steel grill was used. This way if for some reason the experiment explodes between measurements, it will be contained within the grill, several layers of metal, stainless steel, with exhaust vents for any untoward pressure.


  8. Light the charcoal. Be careful, it will spark up at first as the phosphorous in it ignites.
  9. (steps 10-11)

  10. Tear off a strip of the paper towel, lengthwise, fold it over a few times. This will serve as a low-friction gasket between the aluminum can and the tin can, and then wrap it around the aluminum can, then slide the aluminum can with paper-towel gasket into the tin can.
  11. Place tin can on the box, over the lit coals and mark the aluminum can right at the paper-towel line.
  12. (steps 14-15)

  13. Close the safety shield.
  14. Take measurements at regular intervals to determine the heat and the distance the aluminum can is pushed upward.

Data from Single-Coal Materials Test

  • 16:49 – Ignited the coal and set the tin can in place over it.
  • 16:54 – No rise. Tin can warm to the touch.
  • 16:59 – No rise. Tin can hot to the touch.
  • 17:04 – No rise. Tin can too hot to touch.
  • 17:09 – No rise. Tin can too hot to touch.
  • 17:17 – No rise. Tin can too hot to touch.
  • 17:25 – No rise. Tin can too hot to touch.
  • 17:27 – No rise. Tin can too hot to touch.
  • 17:35 – Experiment ends. Temperature of water is 132F/56C. Measuring cup not precise enough to measure negligible loss in water.

Single-Coal experiment end temperature 132F / 56C.

Data from Dual-Coal Materials Test

  • 17:38 – began experiment over from step 5 onward using 2 coals.
  • 17:45 – No rise. Tin can too hot to touch.
  • 17:55 – No rise. Tin can too hot to touch.
  • 18:00 – No rise. Tin can burns to the touch.
  • 18:02 – No rise, but noticed steam coming from between the two cans
  • 18:04 – Boiling noise heard coming from can. Pulled out aluminum can and paper-towel gasket and visually confirmed boiling water.

  • 18:05 – Temperature of boiling water taken: 185F/85C. Can left open, but still left on burner.

Dual-Coal boiling water temperature at 185F / 85C

  • 18:11 – Temperature has lowered to 175 F / 80C
  • 18:22 – Temperature has lowered to 165 F / 74 C

Temperature down to 165F / 74C

  • 18:27 – Final Temperature at 160F / 71C. Called end to experiment.
  • 18:28 – Measured water remaining water at 1/3 cup.

Remaining water at end of experiment: 1/3 cup.


  • The paper towel gasket failure prevented enough pressure from building up to push aluminum can upward. I believe the space-volume of the paper towel itself decreased as it soaked up evaporating water. Either that, or the expansion of the tin-can due to heat increased the gap between it and the aluminum can, which would not have heated at the same rate as the tin can because it was insulated from the heat of the coals.
  • The water in the tin can was gently boiling at a mere 185F / 85C. Using a quick and dirty chart as a reference, this means the internal pressure at the time it boiled would have been about 8-9 psi even with the steam escaping out of part of the paper towel gasket.

Internal pressure of 8-9 psi was achieved.

  • Overall water loss was 1/6 cup, or 33%. This is even after the tin can was opened up to the air for 25 minutes while still over the coals.
  • The coals appeared to begin losing heat at around 45 minutes, though further testing would be necessary to verify this as well as its true burn time.


  • One charcoal tablet in a semi-closed tank will not boil water before fuel burns out.
  • Two charcoal tablets simultaneously burning will boil water in a semi-closed system at a lower temperature than 212 F / 100C due to the difference in air pressure.
  • Internal pressure of the boiler failed to achieve 15psi or greater, thus not enough work was generated to push the aluminum can upward.
  • Water loss of 33% at open top encouraging, but not precise enough as heat dropped too rapidly from open boiler to accurately predict water loss in working conditions. Additional testing necessary.

Final Analysis:

If the engine built requires less than 10psi to run, I should be able to achieve the pressure I need without even brazing the tubes. A rubber lid that doesn’t melt below 200 should easily suffice for the a very low power engine while simultaneously providing a safer boiler for a lower materials cost, because I can just let it continually vent if need be. Though I may not do this, depending on what my other options are. Considering that the coals are designed to burn for approximately one hour each, and the opening for the steam will be considerably smaller, there should be no problems with the boiler tank running dry during the run.

Further Testing Needed:

  • Compressed air-pressure test on tin cans to test pressure tolerance.
  • Water-loss test on test boiler in working conditions.
  • Vapor-pressure test on boiler in working conditions to determine if vapor pressure will exceed pressure tolerance of boiler.
  • Engine test on compressed air to see if it runs at all.
  • Engine test at differing pressure levels to test performance.
16:49 – Ignited the coal and set the tin can in place over it.
16:54 – No rise. Tin can warm to the touch.
16:59 – No rise. Tin can hot to the touch.
17:04 – No rise. Tin can too hot to touch.
17:09 – No rise. Tin can too hot to touch.
17:17 – No rise. Tin can too hot to touch.
17:25 – No rise. Tin can too hot to touch.
17:27 – No rise. Tin can too hot to touch.
17:35 – Experiment ends. Temperature of water is 135 degrees F.

The URI to TrackBack this entry is: https://thesteampunkforge.wordpress.com/2010/05/17/tin-can-boiler-materials-experiment-hard-science/trackback/

RSS feed for comments on this post.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: