Steam Car in a Day

The two machines on the starting line.

Building Frobette

The Design The Shopping List UGLY 
(the car we used)
What we
to build
Early Status Mechanical 
Technobabble Two hours
Race Day
Crew positions 
and roles
Tactics The Race The result Parting Shot

The occasion: An episode of the British, engineering, TV show called Scrapheap Challenge; (In the US, its called Junkyard Wars).

The Challenge: Build a coal fired, 4 passenger, steam race car from stuff scavenged from the scrap pile, then race against the one built by an opposing team.

The time limit for construction: 10 hours.

The Team: Jeff (AKA: dp), the organizer, and author of this treatise. Crash (AKA: Bill) Chief Designer and captain and Geo, lead scrounge. Making a guest appearance was Richard Gibbon. His day job: Head of engineering collections; National Railway Museum, York, UK.Richard Gibbon, our expert.

The place: A corner of a working scrapyard in the Canning Town part of London.

Now this was an unlikely junkyard in one way. It contained three boilers with current inspection certificates, and a like number of engines in varying sizes. Yup, planted, but necessary. Its not possible in 10 hours to build a certifiable boiler large enough to drive a car, if none of the team are UK code certified weldors. In fact, it may not be possible to complete the paperwork needed to certify a boiler of that size, in 10 hours, never mind build the thing.

Note: this document is full of steam engine jargon.  If you don't know your regulator from your Johnson bar, I suggest you take a look at Steam engine basics. This page was written without seeing the broadcast, from notes made contemporary with the build.  If it doesn't match the broadcast, its because they re-arranged reality.

The Design

General theme: As small and as light as will still hold the four of us. The fewer parts the better. We did decide to convert a regular automobile, and not try to build our own chassis. We would be stuck with some compromises, but we decided it would still be a net time-saver.

One of the first difficulties we would face as a result of our decision to use some kind of car as the basis of our machine was that the existing gearing was going to be a poor match to a steam engine. An internal combustion (IC) engine has to be turning at fair speed before any power can be tapped. This means a clutch so it can keep running when not moving and step down gearing between engine and wheels. Typical gearing in a car the size we were thinking of is 4:1 reduction in top gear, and 16:1 in the lowest gear.

By contrast, steam engines hit their red line at a speed usually below that of an idling IC engine. They also can produce full torque from a standstill, so they don't normally use a clutch. So we needed gearing that was in some ways the opposite of that normally found in automobiles.

The vintage steam car
A purpose built steam vehicle often has the engine directly coupled to the drive wheels. (On a railroad engine, they may actually use the wheels and the axle as a part of the crankshaft). If any gearing is present, it will have a final ratio near or even above unity. With the automotive differential typically providing somewhere between 2.5 and 4:1 step down, we would definitely have to do something about this. If we wanted to move any faster than a walk, interposing some sort of step up mechanism was an absolute requirement.

We had a couple of options. Richard's first thought was straight from railway practice - Fit bigger (tractor perhaps?) wheels, and use a direct coupling between the engine and the transmission, drive shaft or even straight to the differential. Our leanings were in the direction of some sort of indirect coupling, one that included some kind of "gearing" between engine and the transmissions input shaft, retaining the normal tires on the donor vehicle. We planned to keep the transmission in place in case our engine didn't have enough power to get us rolling.

Tractor wheel in use

Our coupling possibilities: Assuming we didn't want a direct connection, but wanted a step up, we could use chain (first choice), belts, a differential run backward, a second (or the original) transmission running backwards. Other possibilities included industrial gearboxes (spur gear, not worm gear) or even an automotive overdrive unit (very unlikely to find). At least there was one problem that we wouldn't face: Since a steam engine can produce full torque at a standstill, we could dispense with a clutch.

Other points: To meet boiler safety rules, we had to have two ways to put water into our boiler while it was at pressure. If the engine had "reversing" gear, we might get away without brakes.

Pre-scrounge design summary:

A converted car, with a simple engine; and conventional boiler. Indirect drive retaining the transmission "just in case". As always, what we manage to actually find will affect the result.

The shopping list

What we found

A double acting, single cylinder engine, with reversing gear. Most likely intended for a small steam launch. Looked to be about 3" bore, and 4" stroke. By our calculations, it would make around four hp. Bonus: a displacement lubricator installed. (geo)

Our engine (installed)Outside of our boiler (smokestack clipped)
A view into the boiler, showing tube construction

A small launch boiler, of unusual design. (Counter-rotating tube spirals, top fed, labyrinth fire path.) It even had a smokestack attached. Could have been 10% bigger, but the other ones we saw weighed as much or more than the car we were planning to use. (geo again)

A bunch of pipe and every valve I saw. (dp)

From near the toilets, I found a pile of empty water jugs (as used by the water coolers) (dp)

Steel section: lots of box, a little angle. Some fairly fresh, some rather "experienced. A large piece of light sheet steel, to make a heat shield. (dp)

Tractor wheels (geo)

Rolling stock

The first thing I found, and dragged the rest of the team out to see, came as a real shock to my American team mates. A true British Oddity, in the form of a road tax beating three wheeled "Reliant Regal Saloon". (They are classified as a "motorcycle with sidecar") I had seen one before (in motion no less), they hadn't, apparently not even in pictures. Crash was dubious, he was certain it would roll over in a turn.An intact Reliant Robin.  (not the one we cut up)

In its favor, it was very light, at half our upper limit. (To earn its status as a motorcycle, in addition to only three wheels, the law set a maximum weight of 10 cwt. (1120 lb.. 510 kg))

It had rear wheel drive, and a separate frame. As a bonus, the body was easy-to-remove fiberglass, hence their nickname "The Plastic Pig" (used to help keep the weight down to the motorcycle limit). The brakes didn't work, not even the parking brake. Engine access could have been a problem if we had cared about retaining the original body. There was a little rust in the frame, but we figured it would last the race.

Other choices: A white VW, but it was too banged up (flattened and missing wheels) to consider (no clutch disk for a coupling either). A tiny Fiat was a contender, but it had a steel unit body, and a transaxle, without any sentimental points to offset those negatives. In its favor, it was surprisingly new looking, and the ancillary systems (like brakes) were in good shape.

Status report: The first round of scrounging has happened. We have a likely chassis or two, an engine and steam plant, some steel and plumbing. Still missing are parts to implement the preferred of the possible speed increasing options. (Our second choice, tractor wheels were in hand.) We don't yet have a way to get water into our boiler, but we did have a place to store it. While far from done shopping, this was enough materials for some building to commence (and thus Crash and Richard would have something to do), while more scrounging happened.

Starting to build

We mean it when we say Deconstruction Society

Our robin about to lose its gas tank.  Our engine is in the foreground

We did the sums, and the Reliant got pushed into the shop. The plastic body was going to be a clear time saver. Out came the reciprocating saw with a wood cutting blade, and the circular saw. Fiberglass shards went sailing through the air. We get to use the fire extinguisher; I tried to flame cut the door hinges, not realizing they were aluminum (they are always made from steel here)

So we cut it in half, just above the top of the (tiny) tires, and removed most of the nose, to make access to the engine compartment easier (and to give the driver and motorman an easy escape route, should a steam line break during competition). The 700 cc engine was an easy, two- person no hoist job to haul out. It was leaned against a wall, and someone pulled out the clutch disk, for its input shaft matching splined center section.

A picture showing how much of the body was cut away.

We had found a pair of tractor tires, but decided that they might upset the steering geometry too much. A vehicle with a single front wheel is a very different animal, one that cares a lot about the steering axis angle. My three wheeled vehicles always put the two wheels in front, so while I have seen the equations for delta trike stability, I hadn't ever used them "for real", so there was no chance I remembered them. [I left my copy of Sharp back home] When Geo brought the subject up, I couldn't say just how badly jacking up the back would affect things.  It was clear that it would reduce the trail of the system, and that's usually not a good idea. We returned them to the yard, and the Beach Boys wound up using them. (Richard's idea of using some rubber hose for an engine coupling was also a non-starter. The peak torque on the coupling was going to be roughly the same as Crash or me hanging on the end of a 75 cm pipe wrench.)

I went looking for a step up drive. I dragged in a small motorbike (90 cc), and (as can be expected from me) some bike gearing. Human power curves are not all that different from a steam engine's, and ours was pretty small. Bike chain is plentiful on the heap, so we could find enough duplicate parts so we could double up and be pretty confident of success. At some point a reel type lawnmower appeared, with a whole bunch of chains and sprockets on it.

The motorbikes drive wasn't suitable (wrong ratio); the lawnmower was too light duty. Doubling the bike stuff would mean spacers, and a messy adapter for the larger gears large (110 mm bolt circle) After lunch, I took another pass and discovered a larger (250 cc) off road motorcycle. It had a nice heavy chain with 3:1 ratio sprockets. (Another possibility on our list if I hadn't found the motorbike was the 2:1 chain drive of a camshaft, but the small engines didn't have the typical robust chain drive of the typical American big iron V8.)

Status report: With most of the big and critical parts found and dragged into the shop, and the bodywork out of the way, construction could begin in earnest. The scroungers still needed to find a way to put water into the boiler, but a pause for building happened.  I will start the description with the real heart of our machine, the steam plant.

Surprise! you expected that it would be the engine.  The engine, while full of high precision bits, is not what will determine if our vehicle is a success.  The boiler sets the actual output, and dictates the final performance of the system.  Change boilers to one that works at higher pressure, or with greater capacity, and your machine will be much faster with the same engine.  Fitting a bigger engine without changing the boiler, and your performance might actually decrease.  Likewise, you might be able to use a smaller engine, without a change in performance.  In addition, the boiler requires real skill to operate, especially given the strictly manual controls on the ones we had access to.

Plumbing complexity: While us Yanks were delightfully tearing hunks of bodywork off, Richard was busy with the hacksaw, and the pipe threader. As an expert on British boiler regulations, and having passed an inspection before, we handed him the job of boiler plumbing.

Steam boilers are surrounded with a tangle of pipe and a small forest of valves. Starting at the bottom of the boiler, you will find a water inlet, and the "mud valve" (drain). Our inlet was fitted with a non return valve, and a T fitting for the two water sources. Each water feed had an isolation valve (in our case, scarcity meant the injector isolation valve was manual). Moving up the side, the gauge glass (which shows us how much water was in the boiler, a critical measure) had valves top and bottom, the lower one a three-way valve. A standard part of a fireman's daily routine was to "blow out" the water in the gauge, in order to be sure of an accurate reading.

Photo of boiler fittings, showing mud valve and water feedsPicture of some of the top-of-boiler plumbing, with labels

Moving to the top of the boiler, the tangle continues. The point of the exercise is the outlet pipe (which has a shutoff valve of course). Another boiler outlet led to the pressure gauge and to the safety valve (ours didn't close fully, and hissed until we cleaned it on build day). The main outlet also fed the whistle valve and more plumbing lead to our injector (both a steam valve and a coupler were in the pipe along the way. The water supply to the injector was also provided with a valve).

We aren't finished yet. To improve the fires draft, the steam from the exhaust went into a nozzle mounted in the smokestack (called a blast nozzle). Steam fired up the stack would take air with it, and the fire would burn hotter as a result. To get this helpful effect when the engine isn't running, there is a valved feed from the boiler.  Oh yea, the top of the boiler was also home to the firehole, and the smokestack

A grand total of 13 valves, two accessory fittings (blast nozzle inlet pipe and pressure gauge), and the vent pipe from the safety valve. All this and that's before we talk about the plumbing on the engine itself. All of the pipes exposed to steam had to be rigid iron pipe, with threaded fittings at each joint. Each pipe had to be cut to length, threaded on both ends and the joints sealed with Teflon tape.

The safety valve, since it vented to a vertical pipe, needed a drain at the base. Without the small hole we drilled, water could collect in the elbow, and it would be blown out the top when the safety valve activated. Since that drain would leak steam when the valve was acting (and the vent was right behind Crashes’ head), it needed a deflector. The boiler fittings were finished by fitting a striped card behind the gauge glass (so refraction would make the water level easily visible), and a decorative ornament on the top of the smokestack.

Mechanical details:

Its now after the lunch break status meeting. Our plan was essentially settled. We would put our engine in the passenger foot well, on rails that would let us adjust the chain tension. A throttle would be placed near the engine, in a place where both the driver and motorman could reach it. It meant the steam line passed close to the Johnson bar, so it would need a cover. The boiler was going to go where the rear seat used to be. The feed pump (Geo found one) and main water tank would go in what was the trunk (boot), along with the coal bin. The floor pan was the same freely burning plastic that caused us to reach for the extinguisher when I tried to flame cut the doors off, so Geo fabricated a sheet metal cover to protect it from embers falling out of the fire grate.

The cars original drivetrain was mounted using the typical three-point style. It had motor mounts on each side of the front of the engine, and a single mount at the rear of the engine. Remove the engine, and the transmission is free to flop around. In addition to the problems of getting chain tension, using other than top gear would cause the transmission to spin instead of the drive shaft.

To provide the support that the engine used to, I fabricated a new front transmission mount, and welded it in place. (Not just to the frame, but to the transmission studs as well. Bolts take time, and repair isn't an issue.)  I cut the bell housing away exposing the input shaft to the passenger side. Our motorcycle chain would connect things to the engine, which was going to live in the passenger foot well. Unlike the motorcycle, the big gear went on the engine, and the small on the transmission. Gave us the step up we needed, and left the transmission in place if we weren't happy about the ratio we wound up with.

Annotated picture of the transmission. Includes mount, gearing, and guard

The small, former engine sprocket was welded to the center of the clutch disk by Crash, and slipped (with a little percussive assistance from Geo) over the input shaft. The large ex wheel sprocket was bolted by Geo to a pulley we found that fit the steam engine's shaft (we didn't want to risk messing up the engine's assumed to be carefully balanced cast iron flywheel. Since the pulley was also cast iron, we didn't try to weld it). A ball bearing was mounted on a plate by Richard to support the nose of the transmissions input shaft, and I welded it to the transmission mount after the sprocket got mounted.

A view of the Motormans position

Harvesting technique highlight – Richard and I collaborated to extract a bearing from the side cover of the 90 cc bike I had hauled in as a possible chain donor. He started by unbolting the retaining clips holding the bearing in place. Clips gone, the bearing didn't pop out like it was supposed to, so learning from observation of our more aggressive tactics, he grabbed the hand hacksaw, and started cutting. I wandered by when his cut was part way thru the case. I said, "allow me", took the housing out of the vise, raised it overhead, and WHAM, I applied the (floor or bench I forget) "tool" to it. The housing split cleanly along the partial saw cut, and the bearing popped right out. It fit the transmission shaft close enough.

The engine was carefully lowered onto a couple of pieces of box section. While we did eventually install two bolts between engine and rail, we did a lot of testing with only G clamps to hold it in place (We ran the race with some still in use, and pictures taken from the display at the Donnington steam show have at least one still there). The motorman (me) would have to hang his legs to the outside of the engine, and reach around the steam plumbing to get to the Johnson bar, so a footrest snaked its way out in front of the engine.
We got lucky, and found a nice wire reinforced, flexible, steam rated hose to use for the steam connection between boiler and engine, so we didn't have to worry about engine shaking breaking a pipe. We used ordinary garden hose for the hoses to the water tanks. I think the blast pipe was either a drain hose from a washing machine or one of the radiator hoses from the Fiat. The engine condensate drains were fitted with some fuel line from the Fiat, to keep Crash from getting drenched at startup. (Steam engines, especially older ones can be incontinent)
Rear of car.  Note boiler mounting plates, water pump, and crew seating
As mentioned, the area formerly occupied by the rear seat got a sheet metal cover, and then a second, steel plate on box section spacers, to provide an air gap.  (installed with a lot of muttering by geo) and the boiler was installed smack in the middle (roughly over the differential). The feed water pump was on the passenger side, and needed a cutout in the body to be able to use. Plywood seats were installed over the rear wheels for the fireman and water tender (Richard and Geo respectively), and a grab bar was installed for them to hang onto.

Up front, another footrest happened, this one for the driver (Crash)  along with guards for the chain drive, and a shelf in front for 5 gallon spring water bottles, our reserve supply. (as part of the safety inspection, we cut away what was left of the nose, so Crash had a clear escape route)

Our steam plumbing was completed with my finding one of the injectors hidden in the pile. We now had two ways to fill our boiler, and thus (after inspection) could actually fire our boiler. The steam lines ran between the two seats, and there were valves everywhere (including some tucked behind the seats). We didn't find enough non return valves, so we had to use some manual valves to isolate things like the various boiler feeds. We fitted a cover over the steam hose to reduce the chance of burns. Moreover, it took a couple of tries to get a blast hose to stay on.

Two hours to go: We actually were close to done with the planned construction, with time left. So we hauled in a compressor, and put some air pressure in the boiler, initially to look for leaks. Richard was an excellent plumber; there weren't any of significance. (In fact, the only plumbing problem spots were the compressor hose and the safety valve that didn't want to completely close). We tried our engine for the first time, and it started up, spun to a good speed (no load), and once at speed, we blew our first blast hose connection off. We still cheered.

Status report - Sundown minus two hours:

The hand scrawled name plate - Frobette

We are essentially done with planned construction. Time to do some testing, and see how our machine works. Oh, yea it has a name now, Frobette. After some of our computer geek slang surprised Cathy in the first episode, I presented her with a copy of "The Hackers Dictionary", edited by an old friend of mine named Eric Raymond.  She did some thumbing thru the book, (to decode something one of us said)  she became enamored of the term "frob", and suggested it as a name for our machine. There is a tradition however: steam powered vehicles must have female names, hence "Frobette". I grabbed a permanent marker, and wrote it on the body.

Front of car showing suspension lockout and water bottle rack

With a full head of air, and the sun still in the sky, we did the obvious; Crash sat down behind the wheel, stuck it in gear, and opened the throttle. We had built a steam powered pogo stick, the car jumped significantly with each pulse from the engine. Ok, time to think. The tires were very low, so some of the spring came from that. There wasn't a tire chuck in the tool pile, but Geo knew a trick with the hose, and the tires were filled, without us resorting to hand pumps.

Hard tires helped, but we still had some bobbing. I grabbed a 4x2 timber, held it up to the fully exposed front suspension, drew a line with the grease on my fingertip and cut out a block. Geo made with the sledge hammer.  The timber, some steel wire, a bit of duct tape for luck, and we lost the front suspension. Now we could actually move a little.

Well one lockout was good, how about three. Two more hunks of timber, and we had dispensed with the suspension. We were still getting a bit of spring windup, but not so much, that we felt we needed to add some steel to further constrain the rotation of the rear axle. We had something that acted like a car.

Our only brake
We all took turns getting it to move on air. . There was even time for me to fit a simple brake. As in Flintstone's simple brake – A bit of box section, strap hinged to the front suspension. Retraction was manual, a bit of wire instead of a spring. In addition, it only worked when steering straight ahead. (Our original plan was to use the traction engine practice of shoving the engine to full gear reverse, and gently opening the throttle)

Geo devised a floating cover for the feed water tank, and I put in some hooks to tie the water jugs onto the front of the vehicle. The Sun set and Robert called time. We still had some concerns. While the car had moved under air power, it hadn't moved very well. We rationalized it to the fact that air doesn't have the energy of real steam and that we were only using 6 bar (90 PSI) of air, and we would use 10 bar (150 psi) steam. In addition, of course we had to pass a boiler inspection the next morning, before we could fire up, and see if the problem was just air instead of steam, or something more basic.

The Boiler Inspection: Bright and too early the next morning, we all travel to the build site, to get our work inspected. There were a couple of small problems, but nothing that caused us to fail. Our injector had managed to get itself clogged at some point. We had to take it off, and very carefully clean out the lump of gunk keeping it from working. (By this point, we had fired up the boiler, so it had to be unbolted from some fairly hot pipes before we could clean it up). The inspector asked us to change our gage to one with known good calibration and re-arrange our safety valve vent. We didn't know it then, but our small boiler was going to give us a real advantage, it came up to pressure very quickly. Less than 15 minutes after lighting the fire, we had a full head of steam.

With this head of steam raised, we discovered that our safety valve didn't seat that well. Since any repairs to the valve (and swapping the pressure gauge) would take a cold boiler, we dumped our fire, drained the boiler, and loaded Frobette onto the tow truck that would haul her to the track. We would fix the gauge and valve at the track.
Frobette on the tow truck.  Note strap not attached to steering linkage any more
I managed to avert a disaster, by catching the tow truck operator hooking his hold down strap onto the main steering linkage. A hollow tube "tie rod", it would have been bent when he tightened the strap, and that might have limited how much we could turn in one direction. A worst case outcome; the bumping and bouncing on the ride to the track could have torn it off completely, leaving us with no way to steer at all.

That bit of technobabble wasn't unique to Geo.   He got it from a short film titled Turbo-Encabulator.  No, we don't know where a copy can be found, or who did it.  But I did get him to write it down.

"the drawn reciprocating dingle-arm with sinusoidal repleneration,
mounted on a baseplate of preframulated amulite"


At the site, we made fast work of fixing the safety valve, and swapping pressure gauges. I went to work making kindling, and "well sized" coal. (They provided us with "locomotive" sized lumps. We wanted walnut sized bits, to better match the size of our boiler)

A picture of our safety valve, disassembled to clean and polish the sealing surface

Chopping a block into kindling to start our fire Breaking large lumps into walnut size

The team, tv folk, etc. walking the race route
We got our schedule. Lunch first, then all involved were going to walk the whole racecourse. We would push the machine, boiler cold, over to the pit area, where we would get two hours to raise steam, and make ready. Our small size showed its utility, with an easier push to pit row. Richard provided the requisite oily rag at the bottom of the pile of kindling and we had our ceremonial "first match". We were raising steam for real this time. The wood crackled, and (quoting Crash) "get the wood fire going, then try to put it out by smothering it with coal". Soon, the gauge had started to move, and we could open the "blast valve", and start forcing the draft up the chimney.

Crash lights the first match.  Note safety valve sheild near safety valve The boiler coming into steam

As I hinted above, we didn't realize just what good fortune our small boiler represented. In just under 20 minutes, we were making fine adjustments to the safety valve (Richard was hitting it with the coal scoop). The needle pointed to 150, which meant we had a full head of steam ready, with over 90 minutes until the start of the race. We could use this windfall of time to test our creation, begin to learn its quirks, and even do a bit of practice operation. The build day testing on air had us moving it only about twice the cars length before we ran out of workshop.

The practice came in handy. Steam engines while they can run in either direction, usually have a preference. Our engines preferred direction (especially for starting) was "reverse". The engine wasn't very reliable the times we asked it to start with both engine and transmission set for forward. Once rolling, it was fine, but with a push start racking up a 30 second time penalty, we really wanted it to start on its own.

actually moving under stram Practice start

So, it wanted to start in reverse, but we wanted to go forward? No big deal, we left the transmission there for just this class of problem. Our solution: "reverse-reverse". To get rolling, we would start with the transmission in reverse, and the engine controls in "full gear" reverse. Once rolling we would switch both engine and transmission into a forward gear, in a process of complex coordination that deserved the "fire drill" name it got.

This was a four step process that had to be both done quickly and be coordinated between driver (controller of the shift lever) and motorman (throttle, and Johnson bar). At a signal, the motorman closed the throttle, and started moving the Johnson bar. (This was held from moving by a wing nut, which took pliers to budge). Once the throttle was closed, the driver would start to shift. The shift took time, as the crankshaft (with flywheel) was spinning not just at a different speed (like trying to shift a car without the clutch), but in the opposite direction. You had to use the synchronizer to help bring things to a stop. The motorman would announce that the bar was locked in place, and (in the eventually adopted system) the driver would open the throttle, once the shift was completed. We even managed to shift correctly some of the times during the race. We also blew a few, luckily without any real trauma to our engine. Our coupling system did take some significant knocks while learning to do this. Anything less than the motorcycle chain we chose would have parted.

While steam engines are not the best at containing all their fluids, ours seemed worse than usual. We discovered that the piston gland leaked, as, to a lesser extent, so did the cylinder end covers. Again, with the luxury of testing time, we were able to find the problem and after a bit of tightening, we were significantly reduce (but not eliminate) the leaks. We didn't have to put up with the power (and steam) losses they represented, because we didn't find them until we were actually racing.

Division of Labor:

Unlike a modern car, our adaptation needed more than one person to operate it. This happened partly because of physical layout constraints, you couldn't feed the fire, or pump water from the drivers seat. Frobette could have been operated by two people, but they would have to be well trained, and would be on the busy side, given the "all manual" controls on our boiler and engine. The challenge required the machine to carry the whole team, so we were able to give everyone a lighter workload, and at most once new skill to learn.

How our division of labor applied to the tactics used on the course.  Driving our sort of steam vehicle takes some real planning, as boiler management needs the various players to be anticipating the requirements of course as much as several minutes ahead of our current position. While response to the throttle is as fast or faster than  of an internal combustion engine, the launch boiler we used was intended for use on the water, where rapid change is rarely required, and thus it responded on its own schedule. (It was also a little small given how steep the course turned out to be)

The front seat crew
Starting with the driver – Crash wanted to, so he got the job. Had we not had to perfect the reverse-reverse starting to forward-forward running shift sequence, this might have been the easiest job on offer. With the extra transmission management, it took practice and some awareness of how the steam plant operated. His extra height and strength also came in handy when it was time to operate our primitive braking system.

Sitting behind him was the job with the highest special skill requirement – the fireman. Richard landed this one, as he had the most experience with coal fired propulsion, and the characteristics of the Welsh steam coal they supplied us with (supposedly the best available anywhere). There was one small gap in his experience however. While he has stoked many boilers and was very familiar with the behavior of the specific type of coal provided, he had never fired a boiler like ours.  Its spiraling tubes and top mounted firehole meant it had its own special requirements.

The rear seat crew
We weren't that concerned, he clearly learns very fast. After the build day, he had a pint and a bit of a chat with the other experts. At the track he had me breaking up the big lumps into walnut sized bits, and went around muttering to himself about a "4 inch thick fire", and playing with the firehole cover.

On the other side of the grab rail, Geo manned the pumps, and spun the various valves. His job was actually the most safety critical job on the vehicle. Boiler water level and pressure need constant monitoring. As long as the fire was burning, he couldn't leave the boiler for more than a minute or so. Let the water get too low, and the boiler tubes could get very hot, a panic addition of water in those circumstances could result in a real problem (the kind of "problem" that results in shrapnel, and a crater). Similarly, you didn't want too much water in the boiler; the engine wants vapor, not liquid.

Complicating the job, the apparent level in the gauge glass would change dramatically every time the throttle was opened and closed.  The actual level wouldn't change, the gauge would get confused.  Adding to his concerns: the water he would pump into the boiler was cold by comparison with its contents, he had to add it in small amounts, over time. Adding a lot could all at once would  cool of the contents and cause a serious drop in boiler pressure, something we didn't want to happen while trying to climb the long grade. He had to watch the course, and add water at just the right time, so it could heat while steam demand was low.

The last job open was for a Motorman. This job takes a real understanding  of how steam engines work.  If he were not more needed as fireman, Richard would have  landed in this spot. (in fact, on the Beach Boy's machine, Adrian their expert did take this job) Since I had operated model steam engines in the recent past (admiring my fellow metalworking club members’ creations), I got the job. Operating the engine specific valves, (only 6) and the Johnson bar, I controlled the vehicles speed and, of at least equal importance, the engines steam efficiency.  We didn't have boiler capacity to spare.  If I didn't run the engine for maximum steam economy, we would be forced to stop and wait for pressure to rebuild in the boiler.  Communication with the boiler crew was essential.  It would have been very helpful if I could have seen the pressure gauge.

Unlike a modern car engine with its closed crankcase and pressure oil pump, our steam engine had an open crankshaft.  The motorman gets the oil cans.  The engine requires several different kinds of oil, ordinary (motor oil) for the crank bearings, way oil or grease for the crosshead, and steam oil for the cylinder and valve.

One other detail: You have to warm up a steam engine "gently". You start by clearing the cylinders of condensate.  On the side of the engine are two small valves, that are there specifically there to do this.  You open them, and just barely crack the steam valve, with the engine in neutral. If it doesn't start on its own, give it a small kick.  You want it to turn slowly so steam can flow thru the engine, and it can warm up.  Unlike a car engine, you want to keep your cylinder and plumbing hot.  Steam engines are "lagged", wrapped in insulation to keep them hot, the opposite of an IC engine.


As mentioned above, running our creation took a lot of special skill, and some real planning ahead.  Had we the luxury of a from scratch design, we would have produced a machine that could (within reason) just be driven.  Given what we scavenged, we had to be very attentive to the course, and constantly plan ahead to overcome some of the problem spots in our drivetrain.  Most of our components were intended for marine use, an application that doesn't involve climbing hills, or rapid changes in engine output.  If we needed more output, we had to apply well in advance.

The flash boiler of the antique steam car

Most successful steam road vehicles used a "flash" boiler and liquid fuels. That sort of steam plant is very fast to respond to changes in load. If you needed more power for a hill, you just turned up the burner, and it would make more steam, responding in literally seconds. Our "pot" boiler and coal lump fire couldn't be just turned up and down as needed. If you need a bigger fire now, you had better have added more coal several minutes ago.

Our boiler wasn't large enough to run our engine in full gear all the time. If we were to avoid having to stop and wait for the boiler to rebuild a head of steam when climbing the long grade, I would have to constantly optimize the valve settings. On a railroad or traction engine, the process is called "notching up". It was my job to save some steam for the hardest spots, without being so miserly that it slowed us down unnecessarily.

Since we only had one cylinder, we had to pre-position the engine to get it to start.  We painted a mark on the crankshaft to show us the right spot.  While stopped, one of us would move the engine manually (usually by applying our foot to the flywheel) to get it into the sweet spot.  Full gear reverse would be selected, and when the call for start came, the throttle would be slammed open.  Yes this was hard on the drivetrain, but we had to get rolling for real before then engine hit its dead spot near bottom dead center.  This is the one place where a multi-cylinder (not compound) engine would have helped.

Once rolling, we would shift to a forward gear, and we would proceed on the fairly level flat section of the course.  The engine was easy to trim for minimal steam use, while Geo and Richard got the boiler into peak form for the long climb we would soon be making.

If Geo, Richard, and I did our jobs right, we would start the climb with a perfect head of steam. Crash and I would shift at the bottom of the climb into our best gear for the grade. While we climbed, I would "tune" the engine so it would hold good speed on the climb, while hopefully arriving at the top with somewhere around half the pressure we started with. We would try to get it back up to full pressure on the descent, but we didn't want to start from too deep a hole.

Therefore, on the mostly coasting descent, Crash and I would shift back to a high gear, and I would set the engine so it was doing not much more than coasting. We wanted a little steam to run thru the engine to help the fire along, but this was our time to recover from the work of the climb. Only on the last lap would we try to accelerate on this part of the course. (The start/finish line, and pit row were about ¾ of the way down the hill)

cruising on the downgrade

Richard would take the opportunity to feed and rake the fire, and Geo would try to add, at just the right speed, the perfect amount of water. The goal for all of us, was to arrive at the bottom with a strong fire and a full boiler at its’ optimum temperature. On the flat, we would try to hold onto some of our top gear speed, but the upcoming climb had a priority claim on the steam.

The reason we didn't just coast down the hill, engine off has to do with the "blast pipe". The exhaust from the engine is sent up the smokestack, inducing a forced draft that makes the fire burn hotter and faster. It also kept the engine warm.  Therefore, it was actually a help to run the engine (at minimum settings) instead of coasting down the hill. This draft assist is so important for a good fire, that one valve in the forest of valves atop the boiler, bleeds steam from the boiler into the smokestack. It is needed to help get the best fire when the engine is not running, so you have to remember to open it when you come to a stop. (It is equally important to remember to close the valve when you are rolling again. Once the engine is running, leaving the auxiliary blast open is just wasting steam)

The Race

We pushed up to the starting line, and we put on our fancy helmets.  The race itself was to be a "Le Mans" start (run to the machine), followed by 3 laps of the course we had walked.  We were required to make one stop to pick up water and coal.  Lord Montague was to call the start.

pushing to the starting line

We lined up, and start was called.  We raced to our seats, slammed the transmission into reverse, and opened the throttle.    After a lurch or two, we were off and running, chasing the smooth starting machine of the Beach Boys thru the flat, narrow part of the course, where passing was impossible.

The Beach boys on course
We roll in pursuit of the beach boys

There was some confusion at the tight turn, but we managed to pass their machine by taking to the inside.  We settled in for the long climb, took the left turn at the top, and started to rebuild our resources on the long downhill sweep.  We had elected to pit in the second lap, so we cruised past the entrance.

at speed, aproaching the corner just above the pits

I have no idea of how fast we got up to. While we tried to preserve the speedometer, it  apparently was already dead. The course was far from flat, but with careful notching up, the boiler twice managed the long grade without a stop to rebuild pressure.

We were cruising down the hill, having elected to pit next lap, when the production staff waved us to a stop on the level bit of the course.  It seems that the beach boys engine coupling had broken, and they were stopping the race to let them fix it.  We requested to back up, so we could get the pit out of the way, but the request was denied.  So we stoked our fire, filled the boiler, and set the engine to slow turning, so it would stay warm while we waited.  It took about 10 minutes for them to fix their machine, and we restarted from where we stopped.  We chased them up the hill, and our fast coast downhill had to be sacrificed to the pit stop.  With no tires to change, we were in and out in a very few seconds.  We charged off after the beach boys, as while we were a lap up, we still could be caught, their boiler was large enough to let them climb in full gear, with the safety srill venting excess steam.

Again we caught them on the flat section.  Again we couldn't get past until the turn. (Crash's years as a Boston commuter came in handy when passing in tight spaces).  On the outside this time. Unfortunately, the extra start of the pit, left us a little low on steam.  About halfway up the pressure started to fall.  But they weren't on our tail -- it seems they had broken again.  The report came: 80 psi, and they are pushing!

We were now two laps up, and they had 30 seconds penalty for pushing, we could afford to wait. So we stopped by the side for a "blow up" (letting the boiler come up to pressure).  As we waited for a full head of steam, they pushed past us.  We weren't going to panic.   In fact, we waited until they cleared the climb, so we wouldn't be on their tail.  Pressure up, and the course open, we got under way again.  We came flying down the hill, and with them sitting  in the pit, we put up our arms in victory as we took the tape.

The obvious outcome

The New England Rubbish Deconstruction Society; The NERDS. The first US team to compete in the British Scrapheap Challenge series. (Re-badged as Junkyard Wars when shown in the US)


The NERDS home page

Bog Standard - For Sale