Keep Mediocrity at Bay
Throughout the 20th century, Britain produced many remarkable – in some cases world-changing – internal combustion piston engines. Unfortunately for the everyday motorist on the street, most of them were to be found in aircraft, ships, railway locomotives, motorcycles and even a few racing or luxury cars.
Even in the post-WW2 era, all-iron in-line engines were the staple offering for the British Big Six, side valves were still commonplace, and the RAC horsepower rating system cast a long (stroke) shadow over cylinder proportions even after Britain introduced a flat tax rate. Such engines were not particularly powerful, nor even efficient, but were trusted by the motor trade and buying public, and were forgiving of unsophisticated and imprecise casting, machining and assembly methods.
Into this cautious, conservative, yet somewhat confused land came Palmer’s flat-four, designed and developed in wartime.
The Javelin engine was not revolutionary in its design, but it was far more exotic and eclectic in its influences than anything found in comparable British cars. The aluminium block was formed in two parts, vertically split at the crankshaft centre line. Cast iron wet cylinder liners were used, with an open deck arrangement. The cylinder heads were symmetrical, a clever trick which mean that there was no need for different left and right heads. Aluminium alloy was tried for the heads of some early experimental engines, but cast iron was adopted for production.
The overheard valves are in an in-line configuration, operated through rockers by pushrods and a centrally mounted camshaft on the top side of the engine. Zero-lash hydraulic tappets were specified. Crossflow porting was used, with the exhaust manifolds on the underside of the engine. In all applications each cylinder bank was fed by its own Zenith downdraught carburettor in order to avoid long inlet tracts.
The ‘clean sheet’ design allowed ancillaries to be positioned for easy accessibility in the Javelin – and later Jupiter – engine compartment, and notwithstanding the engine’s complexity, it could be removed and replaced within an hour.
Before he departed from Jowett in July 1949, Gerald Palmer had sounded concerns about the company’s promotion of the Javelin through motor sport. The cars’ successes were the result of the rightness of Palmer’s design; advanced suspension and steering, a low centre of gravity, and a light, low-drag body. Nevertheless he agonised that the engine was still at an early stage of development, and was being stretched beyond the limitations set for its purpose of powering a “utility saloon”.
So were the parent’s anxieties justified? Racing and rallying Javelins and Jupiters were no more blighted by breakdowns and retirements than their peers, but engine failures in service were sufficient to warrant at least eighteen significant engineering improvements to the engine from the start to the end of series production. Crankshaft breakages and cylinder head gasket failures were the best-known issues. The head gasket issue usually resulted from wet-liner shrinkage or sinkage. A copper-asbestos-steel gasket and a re-specified bottom liner seal largely addressed the problem, with greater security provided by the stiffer mid 1952-onwards Series III cylinder block / crankcase castings.
As early as 1950, the hydraulic tappets were replaced with solid pushrods and conventional adjusters. Component supply problems and failures of the adjustment-free tappet mechanism as a result of lubricant contamination were stated as the reasons for the change. Armstrong-Siddeley adopted hydraulic tappets around the same time as Jowett, and reverted to the solid type with even greater haste.
A 1949 change to the hardening process for crankshafts seemed to address problems in early production, but following later failures, a more rigorous investigation by Keighley Laboratories, an external consultancy, identified weakness of the two piece cylinder block / crankcase castings as the primary cause of crankshaft failures. This problem had not manifested itself in development and early production as the first 2000 engine blocks were sand-cast, rather than gravity die-cast, and were stronger as a result.
The improvement process gained pace with the appointment of Donald Bastow as Chief Engineer in mid-1952, effectively eclipsing Charles Grandfield, who left Jowett the following year.
The mid-1952 on Series III engines had additional cast-in ribs in the block sections to increase strength, and broadened oilways. The Series III engine was at that time work in progress insofar as a gruelling testing programme was instigated as soon as the new block was put into production. 40,000 miles of intensive testing at the MIRA test track near Nuneaton on each of three Javelins was completed without any engine failures.
The improvement process continued with further refinement of liner to block sealing in 1953, and in 1954 the forged black-sided crankshaft was introduced with unmachined webs and a new post-forging processing method. Cylinder liner and seal design refinements continued, to address the problems with differential thermal expansion and liner sinkage relative to the block deck resulting from internal forces.
The Holy Grail of Jowett crankshafts was the drop-forged oval web produced by Laystall. This was developed by Dr. Ker Wilson, an engine specialist at De Havilland. His design incorporated refinements to the balance weights to counterbalance rotational oscillation and end-to-end movement.
To their great credit, Jowett’s management and engineers were unrelenting in their work to exorcise the alloy block flat four’s weaknesses and refine it into a reliable and durable engine with above-average durability and reliability. Regrettably the best engines arrived too late, with only the end-of-days Jupiters benefiting from the black-sided crankshaft.
The Jowett flat-four was Gerald Palmer’s only engine design to reach production. Much later in his life he expressed regrets that the pre-1947 taxation system had denied him the freedom to design the engine best suited to Jowett’s purposes:
“I’m not awfully proud of that engine. The unit was designed with the current car taxation in mind, which was based on the cylinder diameter. Had I known this was going to be replaced by a flat rate of tax, I would have designed a short-stroke large-bore engine, which would have been much more compact and lighter, with better combustion and a larger power output”.
Popular Classics – March 1990
The pre-1947 tax system was in essence a tax on cylinder bore, which resulted in engine designs with very undersquare cylinder proportions, and often contrived valve operation and combustion chamber designs which maximised valve area over all other considerations.
As an imperfect principle, long-stroke engines are good for hauling heavy loads from low speeds, as found in light commercial vehicles or overweight saloons. However their high piston speeds are problematic in sustained high-speed use. In pre-motorway Britain this was not a major concern, but in the export or die era, with sights firmly on nation-continents with long-distance highway systems, small capacity long-stroke engines struggled to operate calmly or reliably.
The Javelin’s engine was not absurdly undersquare, and had conventional in-line overhead valves, but the 72.5mm x 90mm proportions compromised future expansion.
The unit was just a little too tightly dimensioned. At a policy review meeting on 20 March 1950 Jowett’s management contemplated stretching Palmer’s engine longitudinally by around two inches to accommodate a 90mm x 90mm cylinder size which would have provided a 2.3 litre capacity, and even the possibility of a very oversquare 1.5 litre engine. I have also been advised that Jowett built an experimental engine with unchanged cylinder centres, and a 3.0″ (76.2mm) bore, giving a 1641cc capacity. Neither idea came anywhere near to production.
Much more recently, a Javelin engine was successfully bored out to accommodate 3.1″ (78.74mm) pistons in overbored standard liners giving a 1755cc capacity.
All talk of expandability is of little consequence. Palmer’s greatest failure was to recognise the limitations of the British engine production infrastructure, and the amount of development effort required to de-bug a “new-right-through” engine. To the credit of his successors at valiant little Jowett, the company rose to the challenge and addressed the problems one by one in a rigorous and methodical manner, and by 1954 had the makings of an outstandingly good engine.
Over the half-century which followed, the British automotive industry produced far too many troublesome engines, and also ruined some good ones. Some were produced in vast numbers, with complete denial of points of failure far more egregious than the Javelin engine’s.
We should regret that Jowett’s conscientious efforts never brought their just reward.
My thanks are due to JCC members John Cash and Geoff McAuley for their valued assistance in providing historical information, and the specialised knowledge of present-day Jowetteers.
Driven To Write 
Thank you, Robertas, for a really good explanation, even for someone like me that, beyond understanding the principles on which an internal combustion engine works, has a limited further technical knowledge.
Do we know why the British government settled on bore size (rather than cubic capacity) as the basis for taxation? Was this done elsewhere? It seems an odd decision.
A very good question Daniel – to which there will doubtless be more than one answer! I suspect that it is a classic case of the government of the day being advised by experts….. in this case, the Royal Automobile Club (not for nothing was it officially known as the RAC Rating). And who exactly were the expert members of this club with the word Royal in its title? The landed gentry, of course; few of whom actually fully understood how any mechanical device functioned. But it’s all simple enough – size always matters so the bigger the hole that the piston fits in, the more power – right? Never mind how far down it goes….
Sorry – leaving aside the flippancy, it was clearly a lack of proper understanding of what was a very new phenomenon balanced against a wish to exploit its tax revenue earning potential.
Indeed, it seems odd. I never knew this.
Thanks for that lucid explanation. In some ways engines are a bit problematic but I can´t help it that the various solutions to managing explosions is so fascinating to me. The engine diagrams are really lovely – the moving parts one is a good extraction of the information. And lastly the funny paradox that England was a hotbed of radicalism in engine design as well as a mire of rutted thinking on the same topic.
As far as I can ascertain England, France, Germany and USA all used formulae based on the diameter of the bore squared. Obviously if one value of the calculation is squared it will have a disproportionate effect on the outcome. I’m still not sure why this seemed to have a greater impact on British engines.
Please rest assured that German car tax always was based on the swept volume of the engine with complete disrespect for bore or stroke (in recent times a factor punishing diesel engines was introduced and a factor for carbon dioxide emission but still based on the volume of the engine). The only curiosity is that the number of π to calculate the piston area from its radius is set at 3.12 instead of 3.14.
French taxation at the time we are looking at here was based on the swept volume of the engine and the numbers of cylinders and a factor for the revolutions the engine could reach.
None of these schemes has a factor looking at the bore of an engine as a single factor.
So the 50-62.5 hp 1.5 engine in the Javelin and Jupiter was potentially capable of displacing as low as 40-50 hp 1185cc up to as high as 77-96 hp 2290cc (rivaling the likes of the Hotchkiss Grégoire), all things considered that is not too bad had Jowett been in a better position in remedying the engine’s flaws (and having a larger concern acquire a stake in the company beforehand).
They could have replaced the old 19-25 hp 1005cc Bradford flat-twin SV engine (that was itself to be superseded by a IOE version) with the 1185cc flat-four.
Regarding the possibility of a very oversquare 1.5-litre version of the flat-four, is it known by how much the oversquare 1.5 unit’s stroke was to be reduced from the existing undersquare 1.5 unit’s 90mm stroke?
Bob – the short stroke 1.5 was a notion of my own, but so was the 90mm x 90mm 2.3 until JTC told me that exactly such a thing was once considered seriously by Jowett’s management. I reckon 90mm x 58mm (1476cc) would work, especially for a racing engine.
Moving on to the RAC horsepower rating, the formula is:
(Cylinder Bore squared x Number of Cylinders) / 2.5
The 2.5 figure is a constant based on three standard assumptions
Mechanical Efficiency = 75%
Mean Effective Pressure = 90lb per square inch
Mean piston speed = 1000 feet per minute
Not by any means an arbitary method, but beneficial to long stroke proportions.
Remember this was from 1910, and there was probably no accurate or consistent method of measuring the power of an internal combustion engine for classification or comparison purposes.
I suspect the physics behind the formula originated in steam engine design. It was important for engineers to have a dependable method of establishing that a one-off engine would deliver sufficient power for its intended task, and you couldn’t put a locomotive on a test brake, or a steamship on a rolling road.
The lowest stroke looked at during development for the flat-four was 78mm for the earlier 69.5mm bore 1184cc engine, which with a 90mm bore would equate to a displacement of 1985cc with a near square 78.74mm bore and 78 stroke equating to 1519cc.
A very interesting article, as usual – thank you.
Re the bore tax, Wikipedia says that the tax was set to favour British engine designs and to slow sales of the Ford Model T, which had a relatively large (and robust) engine. Sounds plausible. Just demonstrates the law of unintended consequences – what was meant to help eventually hobbled the domestic industry.
A few questions come to mind about the Jowett engine – I wonder what the inspiration was for it, given its unusual design. Secondly, it would be interesting to know whether the engineering team ever contacted the government to ask if the tax rules were going to change in the near future.
Lastly, it never fails to surprise / grimly amuse me how a component or set of parts can perform very well in tests, only to misbehave in production (the Hillman Imp being a good example, although I know the testing regime didn’t mimic real life, sufficiently). The Jowett’s crankshaft problems were unfortunate, but they should have tested for the effect of changes made to the production process (he says, from his armchair). No doubt time and resources were against them.
Charles ponders the inspiration for the engine design – I would suggest that there were two deciding factors.
First, it was a clean-sheet design and engineers, given a free choice in such matters, will go for horizontal opposition since it removes several compromises. The result will be more costly to produce (which is why Issigonis wasn’t allowed one for the Minor) and therefore “unusual”.
Second, all production Jowetts, from the very first in 1906, had horizontally opposed engines. Some experiments with in-line prototypes in the late ’30s were not a success and the few examples built were scrapped. Jowett and horizontal opposition were synonymous.
Thanks, JTC. I had no idea – I was wondering if they had seen the Beetle’s engine.
The Jowett story’s incredibly rich for such a small company. I’m getting more and more frustrated that they failed.
Lovely series.
The Beetle flat-four engine was pretty oversquare right from the start, even prewar. Probably not that likely British industry had much chance to look at it then, and air cooling probably made it uninteresting anyway if one has a certain outlook on life. The short stroke allowed the Beetle engine to have short connecting rods and shorter pushrods than Palmer felt he was stuck with on the RAC rating. Too bad. That Laystall crankshaft looks the business, though, I must say.
That the hydraulic lifters didn’t work out is a bit disappointing. They had been around for a while, but the oils available and poor filtration thereof were instrumental in delaying their widespread adoption in the US until the middle 1950s. Adjusting tappets on V8s is a not much fun and time consuming.
Interesting that what goes around comes around again. The quest for efficiency has shown that long-stroke engines are better thermally. So for cooking engines ( and reading the free technical articles Honda publishes), they seem to like about a 0.80 bore to stroke ratio, and Toyota agrees. They put their money where their research leads them, and the cooking motors of both companies and my Mazda are long stroke, as are most VWs.
The Honda 1.5 turbo that powers likely over a million new vehicles worldwide each year these days has a bore and stroke of 73 by 89.5 mm. The Jowett engine was 72.5 by 90 mm. Essentially identical. Makes you think.
A friend has just had an early, hydraulic tappet (lifter) engine rebuilt. It is running beautifully and sounds really good; he makes the point that oil cleanliness is essential, along with strict observance of service intervals – something which owners in period, as ever, could not be relied upon to do. Hopefully this particular engine will benefit from modern oil; the Javelin it powers is being used at every opportunity (but don’t tell the thought police).