A Forgotten American Pioneer:
 
The Thermite Diesel

by Adrian Duncan



Click on images to view larger picture,
hover over the images for a description.
    Background
    The Thermite Diesel Described
    The Thermite 34 Diesel On Test
    Aftermath: The Vivell Diesels
    Conclusion

This time, we stray well off the beaten path to look at an American model diesel engine which has a unique historical significance but which nonetheless appears to have been more or less completely forgotten today, largely by virtue of its extreme rarity. Indeed, we ourselves forgot it in several earlier articles on American diesels! We're speaking of the Thermite diesel, an unusual design which was produced in extremely small numbers in Oakland, California immediately after WW2.

The Thermite diesel goes back to the earliest years of commercial model diesel manufacture. According to the ever-authoritative Engine Collector's Journal, the engine appeared in the latter part of 1945, well over a year before the introduction of any commercial British diesels from the likes of BMP, Owat, Mills Bros, ED, Kemp, Majesco, Clan, Aerol, and FROG. This date is supported by the fact that the Thermite range with which this diesel is associated disappeared around the end of 1945, for reasons which will become apparent as we proceed.

The extreme rarity of the Thermite diesel may well imply that its manufacture was more of a small-scale experimental venture than a full-scale production effort. This does not alter the fact that it was one of the earliest commercially-produced post-war model diesels regardless of country of origin. It was undoubtedly the first American product of this type. Moreover, as we shall see, the engine turns out to have incorporated a quite unexpected degree of original design thinking. Indeed, this model provides ample evidence to justify placing its manufacturer, Jim Brown, at the forefront of pioneering model diesel designers.

My own awareness of this ultra-rare engine was first stimulated by the fact that my valued acquaintance and colleague Ted Enticknap had an example—the only one that I've ever encountered personally. This is the unit that appears as the Thermite 29 diesel on page 222 of Tim Dannel's indispensable American Model Engine Encyclopedia (AMEE), a truly outstanding work from which I gratefully acknowledge having drawn many of the background details which appear in the present article. Following Ted's unfortunate decline in health to the point where he became confined to a nursing home, I was privileged to acquire this very example during the family-organized sale of Ted's amazing collection through SAM 8.

Having set the stage, let's review the background to the development of this highly intriguing model engine.

Background

The model diesel (or more correctly, compression ignition) engine was a wartime European innovation which entered a rapid and very productive development phase in the years immediately following the conclusion of WW2. At the time in question, the commercial establishment of the miniature glow-plug still lay almost three years in the future. Consequently, the early post-war model diesel powerplant was competing only with the well-entrenched spark ignition models for market attention.

The fact that the model diesel did not require a heavy and potentially undependable airborne ignition support system gave it an obvious advantage over its spark ignition counterparts. Naturally, this advantage was just as apparent to American modellers as it was to their European colleagues.

The initial exposure of US modellers to the model diesel came about in 1945 as a result of the early post-war importation of wartime European examples, predominantly French and Italian, by servicemen returning from trans-Atlantic duty in WW2. Interest was further stirred by the publication in the March 1946 issue of Model Airplane News (MAN) of an article on diesels by Jim Watson. It wasn't long before such famous hobby supply houses as Polk's were importing European diesels like the Movo D-2 from Milan, Italy. Polk's initial advertisement for the D-2 appeared in the October 1946 issue of MAN. For context, it's interesting to note that it was around this date that the first commercial diesels were beginning to appear from British manufacturers.

Impressed by the performances put up by the best of these imported European diesels, a number of US manufacturers decided to enter the model diesel market with home-grown designs of their own. It may come as something of a surprise to some readers to learn that the number of American manufacturers involved with model diesel development at this pioneering stage almost certainly exceeded the British total during the same period.

One of those who clearly sat up and took immediate notice of the new technology was Jim Brown of Oakland, California, maker of the Thermite series of model engines and later responsible for manufacturing the Vivell range. Jim was a very experienced model engine manufacturer, having got his start in Sacramento, California several years prior to America's entry into the war. He had begun making model spark ignition engines for his own use in 1937, initiating the commercial production of such motors under the Little Dynamite brand name in 1939. Under the Little Dynamite label, Jim manufactured a series of .374 cuin. crankshaft front rotary valve (FRV) models.

In 1941 Jim relocated to Oakland, California, concurrently switching his brand name from Little Dynamite to the equally explosive Thermite. The Thermite series was launched with an FRV model which was basically a .49 cuin design but was actually offered in several displacements. It later became the prototype for the Vivell "Forty-Niner" sparker of 1946. A .588 cuin. FRV model was also added to the Thermite range in 1941.

Despite the understandable deflection of American public and manufacturing sector attention away from models and modelling during America's war years from December 1941 to mid 1945, the production and further development of the Thermite range nonetheless continued at some level throughout this period. Subsequent models included an improved version of the .588 cuin. FRV model in 1942, followed by a .604 cuin. rear drum valve race car engine in 1944. The latter model was offered in both battery-supported and magneto-equipped versions. It was also the only Thermite model to actually display the name Thermite, which was cast into the bypass passage. Production of the .49 model and its derivatives also continued at a relatively modest rate during these years.

The Thermite engines seem to have been mainly built to custom order at this stage, although a few were reportedly seen in Bay area hobby shops from time to time. As a consequence of this situation, the engines acquired some brand recognition in the Oakland area but were practically unknown elsewhere.

In 1944, Earl Vivell entered the picture, making arrangements for Jim Brown to commence manufacture of the Vivell model engine range. The first such model was the .347 cuin. Vivell "Class C" FRV design which was in fact a derivative of the earlier Comet 35 from Chicago. This was supplanted in 1945 by the Vivell 35, which really set the Vivell range on the road to marketplace success. Jim Brown's involvement with the manufacture of the Vivell engines soon reached the point where he was left with no time to carry on with his own Thermite range, and as of 1946 that range had been consigned to history.

But not before Jim had engaged in what proved to be the first of several forays into the world of model diesels! It's clear that he must have somehow become aware of the model "diesel" (actually compression-ignition) principle very early on, perhaps through wartime correspondence with a serviceman friend in Europe, since within a matter of months following the mid-1945 cessation of hostilities he had produced a few examples of our subject model, the Thermite diesel. The fact that this engine was manufactured under the Thermite name at a time when Jim was becoming increasingly focused on making the Vivell range presumably indicates that at this stage Earl Vivell wasn't interested in diesels.

However, Jim Brown was! Consequently, he somehow managed to set time aside from his commitments to Vivell to make a few examples of this very interesting model diesel engine. In doing so, he became the first US commercial manufacturer to produce a model diesel—more than that, he was among the very first post-war producers of model diesels world-wide, being preceded only by such pioneering European marques as Micron and Jidé in France along with Etha and Dyno in Switzerland which had become established during the war years.

This makes the Thermite diesel an unusually interesting and significant model for study. Apart from being the first American model diesel to see commercial production (however limited), it proved to be the final Thermite model of them all since the pressures of Vivell production claimed Jim Brown completely after 1945. As far as is presently known, it was also the only sideport model ever manufactured by Jim Brown. To cap it all, the engine was probably manufactured for only a month or two at most, making it one of the rarest model diesels of them all.

Let's take full advantage of the present opportunity to have a close look at this very interesting and significant engine.

The Thermite Diesel Described

As received, this example of the Thermite diesel was missing the prop driver and needle but was otherwise in new condition apart from one clearly non-original feature to be discussed below in its place. None of the components bore any marks of previous use apart from some minor scuffing on the backplate inner face, perhaps indicative of a little turning-over by hand.

The rarity of this engine is such that I was unable to find so much as an image of a complete original unit from which to copy the missing parts. Accordingly, the prop driver and needle which appear in the attached images are functional replacements of my own design and manufacture. The needle is a stand-off copy of the type of component often used on the Thermite spark ignition engines, so it's probably not too far off. I have no way of evaluating the accuracy of the repro prop driver—all I can say is that it works well and looks right! I can always make new parts if authoritative information regarding the originals ever becomes available.

When describing an unfamiliar engine, it generally seems best to start off with a few vital statistics. Here we encounter our first major surprise! The Thermite is a sideport compression-ignition engine having bore and stroke measurements of 0.770" (19.56 mm) and 0.730" (18.54 mm) respectively. So far, so good—but these carefully checked (and repeatedly re-checked) measurements yield a calculated displacement of 0.340 cuin. (5.57 cc)! Try as one may, the old abacus stubbornly clings to these numbers! So this example of the Thermite 29 (as it was presented by Tim Dannels) isn't a 29 at all—it's a 34! It's entirely possible that a .299 cuin. variant of the engine did exist (reducing the bore by 50 thou to 0.720" would do it), but this ain't it!

At this point I wish to make it absolutely clear that this in no way represents any error on the part of our good friend and valued colleague Tim Dannels! Tim didn't own this engine, thus being entirely dependent on information supplied by Ted Enticknap for the engine's entry in AMEE as well as an external examination during a visit. Although I very much doubt that Ted ever measured the engine, I know that he truly believed it to be a 29, probably on the basis of first-hand knowledge that such an engine was in fact produced by Jim Brown. Ted was unusually well-informed on such matters—I always considered him to be a reliable source, as did Tim.

The Thermite 34 (as we must now call it) is a very neat and compact design for a diesel of its displacement, weighing in at a commendably light 8.36 ounces (237 gm) in its restored running condition as illustrated. This is actually quite comparable with the later British 5 cc K Vulture Mk. II diesel, which in its day was considered unusually light for a 5 cc diesel at 7.93 ounces (225 gm) and considerably less than such models as the ETA "5" diesel of 5 cc displacement, which tipped the scales at 9.5 ounces (269 gm). The engine is also more compact than both the Vulture and the ETA despite its greater displacement.

The Thermite diesel is built around a main casting which incorporates the main bearing housing, crankcase, beam mounting lugs, bypass passages, induction boss and twin exhaust stacks in a single component. The fact that this casting was very cleanly produced using a permanent mould certainly requires some form of credible explanation given the extremely small production figures apparently achieved.

Although certain Japanese manufacturers are known to have employed die-casting even for very limited-production models, this was very far from usual among American producers. The use of die-casting by an American firm would normally be taken to imply either that the intention was to go for higher production of this engine than was actually achieved (as in the case of the AHC diesel) or that advantage was taken of an existing set of dies to produce the diesel cases. The latter possibility is rendered somewhat unlikely by the fact that (as far as we know) Jim Brown never manufactured a sideport model other than this one at any time. Accordingly, in this instance we may have to look a little deeper for a more persuasive explanation of this matter!

When considering this point, several architectural features appear highly relevant. Firstly, there's the appearance beneath the main bearing of a vestigial portion of an updraft carburettor boss similar to those featured on Jim Brown's earlier Little Dynamite FRV models. Secondly, the exhaust stacks bear a very close similarity to those of the Thermite FRV models, as do the mounting lugs and fore-and-aft cylinder hold-down lugs.

Taken together, these features certainly imply a relationship to an earlier Jim Brown model. However, if the diesel cases were cast from a modified die originally intended for an earlier FRV design, the required modifications would have been quite extensive given the previously-noted fact that Jim Brown is not known to have marketed a sideport spark ignition model under either the Little Dynamite or Thermite brand names at any time.

Pursuing this train of thought, I must confess to a strong feeling that the "modified die" scenario does not provide a fully satisfactory explanation of the use of a permanent mould casting for such a low-production engine. Even the modification of an existing die involves an investment of time and effort that appears to me to be inconsistent with the seeming goal of limited production of what must surely have been viewed as an experimental design. The fact that Jim Brown was heavily committed to the manufacture of the Vivell range as of late 1945 would surely have left him with insufficient spare time to put this level of effort into a trial model which stood well apart from the mainstream Vivell production program.

For this reason, I actually believe it to be far more likely that Jim already had a limited number of previously-produced castings which had been intended for a far earlier Little Dynamite sideport project which had been abandoned at some indeterminate point prior to Jim moving to Oakland and switching his brand name to Thermite. When Jim elected to give compression ignition a try, most likely out of plain curiosity to see how this European innovation worked in practise, what could be more natural than for him to draw upon a batch of cases already on hand as the basis for his initial attempt at diesel design? To me, this makes far more sense than the idea of Jim going to the trouble of creating a new or modified die and having a batch of new castings made.

Of course, if the above scenario is correct, then Jim would have had to create the required die during the pre-war Little Dynamite period to produce the castings in the first place. Unless the die had been lost in the interim, he would thus have been equipped to manufacture a new batch for his 1945 diesel project. However, I personally still find the "existing casting" scenario with its reduced level of effort and financial investment to be rather more likely given the surrounding circumstances. That said, I freely admit that this opinion is no more than speculative in the absence of any direct supporting evidence.

One very intriguing feature of this particular casting is the fact that the mounting lugs are un-drilled. At first glance, this might be taken to imply that the intent was to allow the owner to drill the lugs at a spacing which would allow the engine to become a bolt-in replacement for a previously-used model from another manufacturer. However, it seems to me to be far more likely that this example was never finished at the factory and somehow came into Ted Enticknap's hands in its incomplete condition as received. It may even be an overbored .34 cuin. prototype based on an earlier .299 cuin. model having a smaller bore of, say, 0.720" Ted was unusually well connected with the manufacturing community in his day, consequently owning a number of engines in this prototype category.

Tim Dannels recalls that when Ted provided the information about the Thermite diesel which appears in AMEE, he mentioned something about this example being some kind of experimental or prototype model. This certainly fits with the above observations. Indeed, the terms "experimental" and "prototype" may legitimately have been applicable to the Thermite diesel project as a whole. My sincere thanks to Tim for so generously sharing his recollections with me.

The appearance of the Thermite diesel is somewhat deceptive in certain respects. At first sight, the absence of either a plug or any form of compression adjustment on the cylinder head gives the engine the general appearance of being another of the fixed-compression units that were not uncommon at the time when it was designed. However, things are not always as they first appear! The give-away is that strange and rather awkward-looking arm attached to the front of the main bearing where the timer arm on a spark-ignition engine should be.

It turns out that the Thermite is an early example of the application of the eccentric-bearing compression adjustment system. The main bearing is bored eccentrically (off centre) through the tubular bronze bushing which serves as the main bearing shell. This bushing is a light push fit in the centrally-bored main bearing casting, leaving it well supported but free to rotate within the casting. One end of the previously-mentioned arm is formed into a clamp which secures the arm to a protruding front section of the bushing, allowing it to serve as a lever for rotating the bushing in the case. Since the actual bearing itself is reamed eccentrically in the bushing, the effect of turning the bushing in the main bearing casting is to raise or lower the crankshaft relative to the rest of the engine, in particular the cylinder head. This in turn naturally raises or lowers the compression ratio by altering the elevation of the piston crown relative to the cylinder head at top dead centre. In effect, the working piston does double duty as both the power and contra pistons, using the con-rod as the compression screw!

The main bearing is bored off-center by 0.025" This means that the crankshaft elevation relative to the cylinder head can be adjusted through a range of 0.050", equivalent to a similar range of movement for the contra-piston of a conventional model diesel. Provided the basic geometric ratio is set appropriately, this should be more than enough for normal adjustment purposes. The cylinder can of course be shimmed at the base to effect some measure of downward adjustment of the fixed geometric ratio if considered necessary. Measurements taken from my own example indicate an available geometric compression ratio range of 10:1 up to 18:1, which should be perfectly acceptable for a low-speed engine such as this.

The arm which was fitted to the engine as received was unquestionably non-original. It was quite crudely drilled, hacksawed and filed from a piece of scrap aluminium alloy plate, in stark contrast to the truly excellent workmanship displayed throughout the rest of the engine. It was also poorly designed, being both bulky and ill-proportioned as well as coming far too close to the prop for safe operation. Most importantly, it lacked any means of providing some degree of friction between the arm and the main bearing housing, an essential feature if compression settings are to be held firmly given the inevitable tendency of the bushing to turn with the shaft due to friction drag.

Naturally, I have no idea what the original might have looked like. I made a functional unit which looked far more in keeping with the rest of the engine and also worked in the correct manner. My system employs a turned clamp which grips the protruding portion of the main bearing bushing using a 4-40 screw for tension. The rear of the clamp is recessed to slip loosely over the unmachined main bearing housing. A thick fibre washer is located at the internal step formed by the creation of this recess. This fibre washer bears against the front of the main bearing housing, acting in effect as a "brake pad" to keep the bushing in its set orientation while still allowing compression adjustment during operation. I angled the screw-in steel control arm back at ten degrees to keep the tip well clear of the prop

The degree of friction is easily adjustable. The first step is the removal of the prop and driver. A simple sleeve is then fitted which bears against the front face of the clamp—a 15mm socket works well. After this sleeve has been lightly tightened against the clamp face (using the standard prop nut and washer), the clamp is slackened off from the bearing extension. The sleeve is then progressively tightened against the clamp using the prop nut to compress the fibre washer to the point where the required degree of friction is achieved between the clamp, the fibre washer and the main bearing housing. At this point, the clamp screw is re-tightened to retain the setting. The sleeve is then removed to allow the prop driver to be re-installed. Simple in the extreme, but it works well and is easily re-adjusted when required.

The beauty of the eccentric bearing system for a diesel engine aimed at a marketplace which was accustomed to spark ignition operation is of course that the compression arm functions exactly like the timer arm on a spark ignition unit by controlling the ignition timing. Operation of such an engine would be instantly and comfortably familiar to anyone having experience with spark ignition. An important additional advantage of the use of the clamping system to attach the arm to the bushing is that it allows the arm to be positioned in the most practicable alignment once a running setting has been established. Finally, the system limits the upper compression ratio which can be applied, reducing the potential for over-compression and hydraulic lock in the hands of an inexperienced diesel user.

The downside of the system is that adjustment of the compression ratio in this way also has the effect of altering the cylinder port timing. Advancing the ignition timing by raising the crankshaft to increase the working compression ratio also decreases both the transfer and exhaust periods while at the same time increasing the induction period. However, in a relatively low-speed sideport engine like this one, this is probably not a critical issue. The various moving-liner models like the Speed Demon 30 suffered from the same defect, but they ran OK within the limits of their design parameters.

Another factor which will inevitably be affected by the application of eccentric-bearing compression adjustment is the lateral alignment of the crankshaft centre line relative to the cylinder axis. Engines featuring some degree of lateral crankshaft offset from the cylinder axis are known as desaxe designs. Contrary to a widespread belief, the name is not derived from that of its inventor but rather from the French word "désaxé", which means "off centre".

The advantages of a desaxe layout arise from the fact that by moving the shaft centre-line towards the compression-stroke side of the engine (the left side looking forwards for a model engine designed for normal rotation), the rod swing angle on the power stroke can be appreciably decreased. This results in an increase in the leverage applied by the piston to the crankpin on the power stroke together with concurrently reduced piston side-thrust and associated friction losses. It also promotes a measure of additional piston dwell at the top of the stroke, which can offer combustion benefits in addition to lengthening the angular duration of the power stroke by comparison with the compression stroke.

For all of these reasons, the arrangement is widely employed today in full-sized engine practise—the petrol engine in the groundbreaking Toyota Prius hybrid is a notable example. In addition, the desaxe principle has been quite commonly used in certain model engine ranges, the Fox marque being perhaps the best-known example.

It should be immediately apparent that the rotation of an eccentrically-reamed main bearing will change not only the crankshaft elevation relative to the cylinder head but also the degree to which the crankshaft is laterally displaced from the cylinder axis. Hence changes in compression ratio have the potential to affect the degree to which the engine incorporates the desaxe arrangement.

To take best advantage of the desaxe effect, the bushing on an eccentric-bearing compression ignition engine having a centrally-bored main bearing housing should ideally be installed with the actual bearing bore offset towards the compression stroke side, that is, towards the rear-view left-hand side of an engine intended for normal rotation. Another advantage of this configuration would be that the tendency of the bushing to rotate with the shaft due to friction would be resisted by the counter-rotational couple created by con-rod pressure acting upon the shaft offset.

The problem with this approach is that the compression lever will then function in the opposite direction to the timer arm of a spark-ignition engine—to increase compression, thus advancing the ignition timing, one has to rotate the bushing in the direction of normal engine rotation. Considerations of operational familiarity for spark ignition users might thus suggest that the bushing should actually be fitted with the bearing offset to the right, which will give the reverse of the ideal desaxe geometry but will make the compression arm operate in exactly the same manner as a spark ignition timer arm—increased compression and hence more advanced ignition timing will result from rotating the bushing in the opposite direction to normal engine rotation.

An obvious fix here would be to bore the main bearing casting offset to the left so that the eccentrically-bored main bearing itself would be located on or near the cylinder axis at its point of maximum lateral displacement to the right. Any movement in either direction would then have the effect of increasing the desaxe displacement in the desired orientation. This however was not factored into the design of the Thermite diesel.

In the end, it all depends upon one's priorities—it wouldn't take long to get used to either arrangement. Given the advantages cited above, I elected to set this example up for normal desaxe operation, with the eccentric bearing journal positioned to the left of the engine's axis (viewed from the rear). This of course means that the timing arm operates in the reverse sense from a spark ignition timing arm, but that's fine by me! The orientation of the bushing could of course be easily re-set if desired to achieve the opposite mode of operation.

It's actually both fascinating and quite unexpected to find such an early American diesel using the eccentric-bearing compression adjustment system. Those engine aficionados who remember this relatively uncommon system at all generally associate it with the David-Andersen 2.5 cc model from Norway, the French Ouragan 0.9 cc and 3.36 cc models from Paris, the English Airstar 2.15 cc and M.S. 2.5 cc from Luton and Newcastle respectively and the Clan 0.9 cc model from Coaltown of Weymyss in Fife, Scotland.

Now it gets interesting! All of these engines date from 1946 or later. According to Adrien Maeght's excellent and authoritative book "French Model Motors", the 1946 Ouragan models were based upon an almost-forgotten 5.5 cc prototype produced in 1941 in extremely small numbers (around 12 examples only) by the Fargeas brothers, post-war founders of the Ouragan marque. This appears to have been the world’s first eccentric-compression model diesel, also evidently being the only such design which preceded the Thermite. Both the clearly Fargeas-influenced Airstar model and the Clan design appear to have originated in 1947, while advertisements prove that the MS 2.5cc undoubtedly reached the market in December 1947. Although Jan David-Andersen had begun experimenting with model diesels in 1944 while WW2 was still ongoing, he used conventional vernier compression adjustment in his earliest designs. His well-known 2.5cc eccentric-compression diesel did not actually enter production until 1950.

Suddenly, the Thermite diesel takes on a quite unexpected significance! It clearly pre-dates all of the above post WWII models, which were therefore not available to serve as design influences. The twelve examples of the Fargeas 5.5 cc which are known to have been made in 1941 certainly preceded the design of the Thermite, but it seems highly unlikely that Jim Brown managed to get his hands on one of those twelve engines. Still, it has to be admitted as a possibility, particularly in view of the similar displacements of the two designs. It’s perhaps somewhat more likely that some American serviceman friend of Jim’s saw one of these engines while in Europe and made Jim aware of its major design details. Sadly, the truth of this matter is now lost forever in the mists of time.

Whatever the circumstances surrounding its inception, the Thermite appears in fact to have been the second-ever commercial model diesel engine to use eccentric compression adjustment! It also evidently shares with the Fargeas 5.5 cc model the distinction of being the largest-displacement diesel ever to feature this arrangement. Moreover, the seeming improbability of Jim Brown having somehow gained direct access to one of the twelve Fargeas 5.5 cc models from 1941 makes it difficult to avoid the conclusion that this concept may well have been a Jim Brown original rather than a re-hashing of an idea acquired from some European prototype. A classic case of parallel development, in fact, since it's very doubtful that any European diesel manufacturer ever so much as heard the name Thermite!

Once one realizes that we are dealing here with an engine which was obviously the subject of some highly original thinking, we are no longer all that surprised to find upon closer examination that this is in many ways a very advanced design indeed in the context of its time. For starters, the porting is unusually well developed for a sideport unit. There are two milled exhaust ports placed opposite one another at the sides of the engine and discharging through an opposed pair of stubby exhaust stacks. Even without taking the engine apart, it's possible to see that two pairs of generously-dimensioned milled transfer ports are located fore and aft between the exhaust ports. Unusually by the standards of the day, these ports overlap the exhaust almost completely. The same system that was later to be applied so successfully by the likes of Cox, Holland and Super Tigre, in fact! The transfer ports are supplied with mixture through two very substantial bypass passages formed by internal milling of the main casting fore and aft.

Given that this is a sideport-induction design, the presence of the fore-and-aft transfer ports and bypass passages naturally dictates that the induction port be located at the side between them. The Thermite's milled induction port is positioned directly below the left-hand exhaust opening (looking forward in the direction of flight). This necessitates the positioning of the cast intake boss and brass intake venturi well to the left-hand side of the engine.

The cylinder sleeve, cooling fins and cylinder head are very nicely machined in one piece from a single steel billet. A lot of work went into this component! The bore is of course blind. The cylinder sleeve below the cooling jacket forms a spigot which is a smooth push fit in the inner bore of the upper crankcase. The stiffness of the assembly permits the use of only two screws to attach the cylinder to the crankcase, just as in the later FROG "bicycle-spoke" engines.

More very logical thinking is apparent at this point also. At first sight, one would expect to find two long screws holding the cylinder in place from the top. However, this is not in fact the case. The installation holes are drilled vertically down through the cooling fins at head clearance diameter (5/32 in.) for the 4-40 slot-head hold-down screws. This diameter is carried all the way down to the thicker flange below the cooling fins which serves to locate the cylinder vertically in the case. However, the holes through this flange are only drilled to clear the 4-40 threads of the two installation screws. The screw heads thus bear directly upon the cylinder location flange, accordingly having extremely short stems. The effect of this arrangement is to eliminate any hold-down stresses from the working length of the cylinder bore—a very desirable provision.

An offsetting design weakness becomes apparent at this point. For one thing, the use of only two 4-40 screws to retain the cylinder appears rather on the minimal side for a diesel engine of this displacement. This is compounded by the fact that the thread length provided in the case for the two installation screws is pretty marginal, especially since the threads are cut directly into the relatively soft aluminium alloy of the case. The presence of the two bypass passages leaves insufficient material to permit the deepening of the relevant holes, since they already penetrate into the passages at the top. Accordingly, great care is required when re-assembling this engine to avoid over-tightening and possible stripping of these threads. Tight enough and no more is definitely the correct approach here, along with regular re-checks of tightness between runs.

So much for what can be gathered from an external examination of the engine. To learn more, we have to remove the turned aluminium alloy screw-in backplate, which seals to the crankcase with a gasket. Upon doing so, we immediately encounter another interesting design feature. The backplate thread is large enough that a full-diameter full-depth backplate would seriously obscure the rear bypass passage in addition to fouling the piston skirt at bottom dead centre. To get around this, the designer has reduced the diameter of the backplate beyond the threaded portion and has also cut a deep channel into the reduced-diameter section. This maintains unimpeded gas access to the rear bypass passage regardless of the orientation of the backplate when tightened, at the expense of some increase in crankcase volume. It also provides ample piston skirt clearance.

Following removal of the two hold-down screws, the cylinder slides easily out of the location bore in the crankcase, in which it is a very close but smooth push fit. This fit is extremely important in this design because the cylinder location flange is above all of the ports. Hence the crankcase seal is entirely dependent upon the close fitting of the cylinder in the case. Despite this, a gasket is used between the cylinder and the upper surface of the case, even though such a provision seems a bit redundant.

The lapped cast-iron piston is very well made and fitted indeed. It is extensively machined away internally to lighten, although quite substantial piston bosses are retained. Despite the machining efforts, the piston still weighs a very hefty 15 gm—high revs need not apply! The piston fit on this apparently unused example is in fact a bit on the tight side. The 3/16" dia. gudgeon pin is of tubular steel with brass end pads. The sturdy steel con-rod is very nicely machined from the solid. It is very closely fitted at both ends, with almost no discernable play.

Con-rod clearances are quite tight, especially at higher compression settings with the crankshaft raised in the case. This necessitates the provision of a pair of chamfered scallops on the lower edge of the cylinder liner, one on each side. These may be clearly seen in the image of the detached cylinder included earlier in this article.

The one-piece solid steel crankshaft is both cleanly machined and nicely finished. The web is a plain disc with no counterbalance features—a bit odd in view of the very heavy piston. Crankpin diameter is a somewhat skinny 3/16"—a larger diameter would inspire more confidence. The 5/16" dia. main crankshaft journal is a very close but smooth fit in the eccentrically-bored bronze main bearing bushing. The journal diameter seems a bit marginal for a diesel engine of this displacement, but it must be recalled that the shaft is not weakened by the presence of induction passages and is also very well supported in its closely-fitted bearing. Given the relatively low operating speeds and power output which may be expected with this unit, the shaft would probably stand up OK in service with careful handling.

The front of the crankshaft has a square-section segment immediately forward of the main journal to serve as a means of locking the prop driver to the shaft. The length of this square feature played a major part in determining the thickness of the replacement steel prop driver that I made for the engine. The prop is secured in the usual manner using a simple nut-and-washer combination on a 1/4-28 thread.

The engine is completed by the brass venturi tube and spraybar-type needle valve. Oddly enough, the venturi tube is not threaded for installation in the usual manner but is simply a tight push-fit into a barely detectable taper in the cast-in boss on the crankcase. Although a thread and lock-nut would unquestionably create more confidence, the fit is such that no difficulties with loosening of the venturi during running can be anticipated. The venturi has a throat diameter of 0.240 in., while spraybar diameter is 0.125" Overall, this is an unusually free-breathing design for a 1945 sideport diesel.

The needle itself was missing as received, but the spraybar shows that it was of the externally-threaded variety, like those of other Thermite models. It was also quite clear that it had used a coil spring to tension the needle, once again along standard Thermite lines. I made a functional spring-tensioned replacement which bears a general similarity to those used in other Jim Brown designs. There's no way of knowing for sure if the engine ever had a hang-tank, but none of the other Thermite models had tanks. Accordingly, I didn't bother to make up a tank for this engine.

Overall, the standard of workmanship is extremely high. In fact, we have to conclude that this is a very well thought-out design indeed, and one which was most competently constructed by any measure. Given these observations, one might logically expect this engine to perform quite strongly by 1945 model diesel standards. So how does it actually run? Only one way to find out!

The Thermite 34 Diesel On Test

Testing an engine of this vintage and rarity is always a bit fraught, particularly when dealing with a pioneering diesel dating from a time when the need to construct such units to withstand the far higher stresses imposed by diesel operation was not yet fully appreciated. However, I felt that if I didn't give it a go for the record, no-one else would! So, armed with a few suitable test props plus some standard diesel fuel which had proved extremely satisfactory in other engines, I set the Thermite up in the test stand and prepared to boldly go where few have gone before ...

In view of the seemingly rather marginal dimensions of the crankshaft journal and rod bearings, I resolved in advance that I would not push the old Thermite in any way and would back off at the first sign of undue stress on the engine. I also planned to keep running time to the bare minimum necessary to form a valid impression of the engine's performance characteristics—a lengthy running-in period was never on the cards. Hence what follows cannot be viewed as anything approaching a comprehensive or necessarily conclusive test.

The major challenge with larger diesels from the vintage era is always achieving that initial start. The most problematic issue is the tendency for the very high compression resistance coupled with internal friction to "brake" the engine, thus discouraging it from carrying over from the initial firing stroke to the next. The result can be a series of pops and bangs with no actual start—extremely frustrating! For this reason, it's always wise to start off testing such an engine using a heavy large-diameter airscrew having plenty of flywheel inertia. I elected to go with a 12x6 APC prop, a very hefty club which I had found to be highly effective for initial starts with the 5 cc Vulture, and Drone diesel.

I began with the compression lever set for the lowest-possible compression ratio. This is always the safest approach, particularly with a rare and irreplaceable engine of uncertain mechanical integrity. The needle was set relatively lean, as determined by ear using the old squeeze-bottle blowing trick—better this than risk flooding the beast and inducing a potential hydraulic lock.

To ensure that there was fuel in the cylinder, I administered a very light prime before flexing the old flicking finger. Compression seal was outstanding, thus offering no excuse for any starting problems. As usual with larger diesels, the Thermite proved to require a fairly high compression ratio for starting. A healthy crank of the compression lever in the direction of rotation soon had it popping and banging, then making a few short bursts. However, I couldn't get it to keep running until I opened the needle some more. I did this in small increments, finally achieving a start at just under two turns.

Once the engine was running, I quickly had to back off the compression considerably as it warmed up. This is actually a very typical characteristic of large low-speed vintage diesels—they seem to like a lot of compression for starting and then require a significant back-off as they warm up. A fraction less than two turns of the needle was found to be a very consistent starting setting, with best running being established at around 1-3/4 turns. Suction appeared to be excellent.

Running qualities proved to be outstanding. Since the piston fit actually feels relatively tight, clearly having had little or no previous running time, I kept the needle a fraction rich and the compression just a hair lower than optimum. The oil colour quickly cleared up to a nice honey brown, indicating that all was well internally, and the engine settled down to a rock-steady note with no trace of a mis-fire, which it would seemingly maintain indefinitely at the settings which I used. Vibration levels were relatively high without being excessive, although I wouldn't want to push the engine much faster.

The compression lever proved to be both effective and convenient in use—I could easily get used to this system! It appears that the basic geometry of the engine was very well chosen, since the best running compression setting proved to be exactly mid-way in the available compression range, that is with the eccentric main bearing bore set at its maximum left-hand displacement viewed from the rear. The engine was thus operating in desaxe mode with a 0.025" crankshaft offset.

The friction arrangements incorporated into my replacement compression adjustment system proved to have been very necessary. After a few runs, the system was in need of re-adjustment since the level of friction slackened off to the point where the compression control began to creep in the direction of engine rotation, thus inducing a slow but potentially damaging upward drift in the compression setting. Re-adjustment was a simple matter using the approach set out earlier, after which no more difficulty was experienced. Presumably the fibre washer took a "set" over the first few operating cycles, as one might expect. I would anticipate having to re-tension the system fairly regularly when the engine was in service in a model.

Optimum compression settings were very easily established since the control system proved to be highly effective in use, with excellent response from the engine. The needle too proved to be both responsive and non-critical, making establishment of the desired mixture very straightforward. The spring used for needle tension held the settings perfectly.

As expected, I encountered no trouble with the unthreaded venturi tube coming loose during operation. However, the two cylinder hold-down screws did need re-tightening after a few runs. The fact that the very short thread length precludes tightening them down really hard means that periodic checks on their tightness are definitely necessary.

I put on 20 minutes in 4 minute runs, allowing complete cooling between runs following a few test hot restarts. In the latter regard, one thing that was very noticeable was the engine's tendency to run unusually hot. I would hazard a guess that the rather tight piston fit had a lot to do with this—even after the above amount of running, the fit remained very much on the sticky side. A lot more running would doubtless free things up, but I was not prepared to subject this ultra-rare and seemingly irreplaceable engine to the associated wear and tear. Even so, running remained very steady at all times.

Hot re-starts were immediate at running settings with just a couple of choked flicks as preliminaries. Cold re-starts were found to be quite straightforward with several choked flicks and a modest increase in the compression setting. Following the initial start, priming proved to be unnecessary either hot or cold. The engine always started with just a few flicks after choking.

One point that I did check was the Thermite's response to a lighter prop. Given the well-earned reputation of larger diesels for requiring a heavy prop with plenty of flywheel, I thought that it might be instructive to try an 11x6 wood prop just to see what would happen. This far lighter prop was found to lack the flywheel effect necessary to carry the engine over from the initial firing stroke, resulting in a series of pops and bangs but no start, try as I would. A switch back to the heavier APC glass fibre airscrew quickly restored happiness, with an immediate start being readily obtained.

Feeling by this time that I'd subjected the poor old Thermite to as much abuse as it deserved after all these years, I contented myself with checking the available speeds on a small range of seemingly suitable props, keeping the fully opened-out check runs to very short duration in recognition of the residual tightness in the piston/cylinder fit. The still-tight engine got the test 12x6 APC up to a steady 6,000 rpm, which equates to around 0.177 BHP—hardy an eye-opening performance for a 5.6 cc engine, although a lot of air was being shifted. A switch to an 11x6 APC didn't produce as much of an improvement as I was expecting—the engine could only manage 6,300 rpm on this prop, implying an output of around 0.142 BHP. The peak had clearly been passed by the time this speed was reached, with the power curve declining sharply in the characteristic manner for sideport designs.

Tests on a couple of other props yielded 5,900 rpm on an APC 12x7 and 5,400 rpm on an APC 12x8, implying outputs at those speeds of around 0.180 BHP and 0.157 BHP respectively. Taken together, these figures suggest a well-defined peak at around 5,900 rpm, where the output is of the order of 0.180 BHP, with a sharp decline thereafter. The engine was evidently running at its peak on the bench using the 12x7, which would thus appear to be the ideal bench testing prop. Allowing for airborne pickup, a 12x8 would likely have been the best prop to use for flying purposes.

Before we dismiss the poor old Thermite on the basis of these figures, it's important to recall two factors. One, we're speaking about a sideport diesel dating from 1945, when diesel design was still in its infancy, particularly in America. Two, the piston fit in this example is still what I would objectively classify as unduly sticky. With a fair bit more running, I have little doubt that all of these figures would improve quite considerably.

But in any case, the numbers aren't by any means dismal by the standards of 1945 or even those of a few years later—the highly-touted British ETA "5" sideport diesel (which was considerably heavier and bulkier than the Thermite despite its smaller displacement) could only manage 0.181 BHP at 6,250 rpm when tested in September 1948 by Lawrence Sparey, all of three years after the manufacture of the Thermite. I have no doubt at all that a well run-in Thermite could beat those figures. Looked at from this perspective, this must be considered a very worthy first effort at diesel design and manufacture on the part of Jim Brown!

The engine came through its 30 minutes of fame with colours flying high, seemingly ready to do more work if desired. If fully run-in, there's little doubt that performance would improve quite measurably. However, I'll leave that to some future owner—for me, the fun of restoring this rarity to running condition and actually experiencing it in operation has provided more than sufficient satisfaction to justify my acquisition!

I would summarize the Thermite 34 as a well-made, easy-starting and smooth-running engine which is extremely straightforward to manage once running and swings a substantial airscrew at quite sufficient speeds for most sport-flying requirements by the standards of its day. It is also unusually compact for its displacement. Any contemporary modeller who saw one of these in the hands of a competent operator would surely have been quite impressed! Too bad that Jim Brown was forced to abandon his promising diesel project so early. However, there was more to come ... read on!

Aftermath: The Vivell Diesels

It appears highly probable that production of the Thermite diesel did not extend much if at all past the end of 1945, being confined to one or two very small batches. According to Tim Dannels, a number of Jim Brown's designs were made in extremely small batches during the Thermite years, this being one of them.

The very short production life of the Thermite diesel almost certainly had nothing to do with any fundamental shortcomings with the engine itself. Rather, it was likely down to the simple fact that as 1946 rolled around the pressure on Jim Brown to maintain production of the Vivell engines left him no spare time to continue with the Thermite range.

Consequently, the Thermite diesel enjoyed only a very brief production life. The extreme scarcity of surviving examples (the illustrated engine is presently the sole example of which I am personally aware) makes it appear more than likely that total production barely reached double figures. The engine certainly escaped the attention of contemporary model engine commentators.

However, this didn't stop other American manufacturers from pursuing the model diesel concept. First among these were Barney Snyder and William (Bill) A. Ruff, who designed and manufactured the excellent CIE 10 diesel under the auspices of the Compression Ignition Engines Division of Modelcraft Hobbies in Los Angeles. This very capably-designed, well-executed and fine-running engine appeared on the market in late 1946. Geographic considerations make it appear not unlikely that exposure to the Thermite diesel may have played a part in engaging the interest of Snyder and Ruff in the development of model diesels, although the Thermite doesn't appear to have influenced them at all in design terms.

The C.I.E. was quickly followed by other American diesel designs, the most successful of which was the 5 cc Drone fixed-compression model which was sold in two distinct variants. Other contemporary American diesels included the Micro, Air-O, Delong, Speed Demon and Mite designs, not forgetting the infamous Deezil. There was even an after-market variable-compression diesel head for the Arden .099 model. A number of these models possessed considerable merit—even the original version of the Deezil appears to have been a very acceptable design, as we have demonstrated elsewhere.

The EDCO company of Del Mar, California, also developed a diesel model, although this never progressed past the prototype stage. The same applies to America's Hobby Centre (AHC) of New York, whose own diesel development program was also halted at the prototype stage, as detailed elsewhere. A few .604 cuin. Ken diesels were also produced by the Kencraft company of Garden Grove, California. As we stated at the outset, the number of American firms having some level of involvement with model diesels at this time was at least comparable to the number of such firms active in Britain, a fact which may surprise many of our more traditional diesel-minded readers.

It was in this climate of heightened interest in diesels among American manufacturers that Jim Brown got his second chance to work on diesel development. Earl Vivell eventually became convinced that there might be a future for the model diesel in America, and in 1947 (still prior to the commercial introduction of the miniature glow-plug, remember) he authorized Jim Brown to develop and manufacture a series of diesel models under the Vivell name.

The first of these offerings was the 1947 Vivell Precision 10 diesel, a neat and compact .102 cuin. (1.67cc) fixed compression model. This very well-made engine used sand-castings throughout and featured disc rear rotary valve (RRV) induction along with a screw-in front housing. It also used the same twin-port cylinder arrangement which had first been applied to the Thermite by Jim Brown. The engine performed well on a correctly-matched fuel and prop combination.

In 1948, a refined version of this engine appeared. This featured a head having deeper fins and incorporating a contra-piston inside the head (as opposed to inside the bore). Compression adjustment was controlled by an Allen-head screw. The variable-compression feature made this variant far more flexible as regards the range of fuels and props which could be used. The Allen-head compression adjustment was a pain, however—a conventional tommy-bar arrangement would have been greatly preferable.

1948 also saw the appearance of a smaller variable-compression diesel model of basically similar design designated the Vivell Precision 035. This very compact and well-made .035 cuin. (0.57cc) unit did not use any castings, being machined entirely from bar stock. Although also an RRV design, it was set up exclusively for radial mounting direct to the rear of the crankcase, a configuration which actually posed significant problems for the user in arranging an appropriate mounting in a model since the mounting bolts had to be tightened from the rear, making the use of a separate mounting ring more or less mandatory. The earliest versions of this engine had the contra-piston conventionally located in the cylinder bore, but the illustrated later variant reverted to the contra-piston-in-head approach used in the second model of the .102 cuin. design.

The Vivell diesels were doubtless developed to fill a contemporary niche in the model engine market, namely that for lightweight engines of around .099 cuin. displacement or less. Although the Atom spark ignition engines of .098 cuin. displacement had proved both popular and practical, as had the FROG 175 sparker in England, it was generally agreed that this was about the lowest commercially-viable displacement to which a practicable spark ignition model engine could be built, Ray Arden's amazing one-off sub-miniatures notwithstanding. The limiting factor was of course the parasitic extra weight of the ignition support system required for spark ignition, which remained much the same regardless of the engine's displacement. Consequently, the use of spark ignition resulted in dramatically reduced effective power-to-weight ratios as displacements went down. The model diesel dispensed with this extra weigh and also avoided the need to accommodate a plug, thus making its construction and use practicable in far smaller displacements. This was the niche which the Vivell diesels were evidently intended to fill.

The commercial introduction of the miniature glow-plug in early 1948 changed all that, of course. Suddenly the diesel lost its weight and simplicity advantage, having to compete now with the glow-plug engine with its similar absence of airborne ignition support and its generally lighter weight for a given displacement. The American aeromodelling community bought into this concept right from the start, sending the model diesel in America into a slump from which it never really recovered, the later very worthy efforts of McCoy and OK notwithstanding. Glow-plug versions of both the Vivell 10 and 035 models appeared in 1949, but the diesels themselves were history. The Vivell range itself failed to keep pace with the changing times and had disappeared from the scene by 1952.

Conclusion

We hope that you've enjoyed this look at a model diesel engine which possesses an unusual degree of technical interest and in addition can lay a very fair claim to being one of the very first post-war diesel designs of them all. It was certainly the first such product from an American manufacturer. The Thermite diesel barely made a ripple in the marketplace, but that had far more to do with surrounding circumstances than with any evident shortcomings in the engine itself. It was surely instrumental in stimulating interest in diesels from a number of other American manufacturers, at the same time doubtless providing its constructor Jim Brown with valuable experience which stood him in good stead when the time came to develop the later Vivell diesel models.

An interesting engine in its own right, but also one with a substantial legacy. We trust you'll agree that the Thermite diesel and its creator Jim Brown both deserve an honoured place in the history of the model diesel!