Wednesday, April 30, 2014


A lusty 1000cc V-twin from 1924, with 8 valves up top, in an Anglo-American hybrid
It was clear from the earliest days of 4-stroke engine design that multiple valves in a cylinder head had clear advantages over just two; the valves themselves would be lighter, making an easier life for valve train components and valves less likely to break.  It's also possible to move more air through two (or more) small valves than one big one, as the total surface area of multi-valves could be larger than a single valve port, without risking a crack across the cylinder head from a weak structure with one mighty hole.
The granddaddy of all 8-Valve motorcycles, the 1914 Peugeot 500M 500cc parallel-twin 8-V DOHC racer (see story here)
Thus, in the 'Teens and '20s a lot of factories experimented with 4-valve single cylinder or '8-valve' V-twins, especially in the racing world.  Indian was first with an 8-v twin racer in 1912, followed soon after by H-D, while across the pond, Triumph and Rudge were building 4-valve single-cylinder bikes, while Anzani had an 8-v twin.
1924 McEvoy 8-V V-twin racer with Anzani engine, as seen at Vintage-Revival Montlhery in 2013 (see story here)
Today, original 8-valve V-twins from the Vintage era are pretty near the top of the collectible heap, but lovers of performance, and hot-rod Vintage motorcycles, still experiment with installing modern reproduction 8-v cylinder heads onto 'J' series Harley crankcases (as per my post on Harry Hacker's Harleys), and the same with Indian products.  A few have experimented with Rudge 4-v cylinder heads atop JAP crankcases, which sounds fun too, although all these experiment suffer from the same afflictions of the original 1920s designs; inadequate top-end lubrication, and a strain on components due to a sudden and significant increase in power!
One of Harry Hacker's compelling experiments adding a pair or reproduction 8-V cylinder heads to a Harley JD bottom end. (see story here)
In their experiments, today's tinkerers are hardly alone, nor are they the first garagistes to DIY an 8-v engine.  Way back in 1924, it seems the Excelsior importer for Belgium, a Mr Taymans, decided to fit a pair of Triumph 'Ricardo' 4-v cylinder barrels and heads atop an American Excelsior V-twin, making a very handsome road-going OHV roadster, the 'American-Excelsior-Triumph'.  According to the Motor Cycle magazine, he built several of these beasts, although this article is the only evidence I've seen of one...have any survived?

From 'The Motor Cycle', July 24th, 1924:

An American V-twin Fitted with British Four-valve Cylinders

Something new in ‘hybrids’ has been evolved by Mr. R. Taymans, a well-known motor cyclist and motor cycle agent of Brussels.
Agent for the American Excelsior, he has a great admiration for the strength, rigidity, and excellent steering qualities of this machine; he has also an equal admiration for the productions of Britain.  So he has manufactured an eight-valve American Excelsior, employing two four-valve 500cc Triumph cylinders adapted to the Excelsior crank case.
A Triumph Ricardo with 4-valve cylinder head, produced from 1921-28
Standard Parts.

With the exception of a slight alteration in the cams to produce greater efficiency, entirely standard parts are used, and the only structural alteration has been the dropping of the engine almost two inches in the frame.  The standard Schebler carburetor is fitted, and with it the machine will do 78mph; this is increased to 82mph with a three-jet Binks.
According to the constructor, the acceleration is terrific.  Altogether, the machine has been on the road for a full year, and with a sidecar.  It is not purely an experimental machine, but is actually on the market, many of them having already been sold all over the continent of Europe. Complete with electrical equipment, the machine is priced at £132.  Mr. Tayman’s firm is Taymans Fréres, 641, Chausée de Waterloo, Brussels, Belgium. "
A much later Triumph with 8-V cylinder heads - the prototype for the TSX model, with special Weslake cylinder heads...a parallel twin like the 1914 Peugeot, but even 60 years later, the French machine's DOHC spec was too advanced for Triumph!  (read more here)
A pair of c.1912 Indian 8-V cylinder heads, offered on eBay of all places, several years ago (read the story here).
A 1929 Harley DAR 8-valve racer from the Wheels Thru Time Museum, seen at the Pebble Beach Concours in 2010 (read more here)
A 'Teens or early '20s shot of an 8-Valve Indian racing engine installed in a Chief roadster chassis, complete with passenger seat and bedroll up front.  The dream of many a motorcyclist - a high-tech racing engine for the road...


Jason Cormier said...

Everything old is new again, and everything new is old again. It's remarkable how much innovation existed in the early decades of motorcycle design before conservative notions of what should and should not constitute a "proper" motorcycle were set into stone.

It's a common theme I often run into when I'm working on my own articles, and a sad testament to how much backwards tradition holds back first-principle innovation in modern production machines (despite their performance and complexity, they still keep rehashing the same formulas established 50-60 years ago).

I'm sure I've read some work by Kevin Cameron that elaborated on the strengths and weaknesses of these early 4V designs, as I recall their combustion efficiency wasn't great enough to justify the extra complexity, something that wasn't aided by the large included valve angles of the day that were shared with 2V designs (though that Indian head seems to be going in the right direction). It wasn't until the Cosworth DFV established the pent-roof shallow combustion chamber as the most effective layout for a 4V head that things really took off.

The Vintagent said...

Hi Jason,

I don't think complexity or inadequate combustion efficiency was the issue with early 4-Valve designs, as they set many speed records and were clearly faster than their opposition in the 'Teens. A pushrod 4v design is pretty simple with parallel valves (Rudge developed a very clever system for radial valves too).

The delicacy of those Indian cylinders heads in the photo above speaks volumes - not enough finning to cool them, and no positive lubrication at all.

Rudge carried on with 4v cylinder heads, proving their efficiency by winning the Isle of Man TT and Ulster races in the late '20s, and were the last pushrod-engined bikes to win the Senior TT; after that, it was OHC all the time.

Early OHC engines also suffered from a serious lack of lubrication. Kevin Cameron once asked me how Cyclone cams got oil, so I asked friends who own them, and found they use a dip system into oil from a small reservoir, which is refilled by a lidded cup above the cambox. In other words, they weren't lubricated much at all! From which their long-distance racing troubles make sense.

It took Velocette to add positive lubrication from an oil pump to their OHC motors in 1925, which was soon copied by Norton, who knew a good thing when it beat them. I owned a 1928 OHC Humber which still relied on a drip-feed system...wholly inadequate!

occhiolungo said...

Much has already been said of the 4V developments, so feel free to skip the following. :) I wouldn't say that the factories abandoned 4V head due to poor performance or combustion efficiency. The American marques simply didn't need to spend the money on 4V heads when sidevalves (with big motor displacements) gave enough power and cost a lot less to produce. Rudge and the others had good pentroof layouts in the mid20s, and the performance was more than adequate.

Lubrication issues of 4V OHV heads was no different than 2V OHV heads, and the occasional squirt of oil or grease to the rockers was all that was provided. Lubrication of OHC was a completely different matter, as the cam lobes themselves needed positive oiling (see Velocette). But the rockers and valves were still unenclcosed, out in the breeze and lubricated by the rider.

Then and now, performance-based buyers wanted to spend their money on the hottest specification. 4V heads allowed for that, even if the poor Ricardo Triumph was so de-tuned from Sir Harry's specification as to be just a shell of his prototype motor. A real sheep in wolf's clothing. But the buyer got his 4V head. :)

Looking for faults of the early technology is fun and simple. Just a quick glance at many of the rockers will show some terribly weak and flexible designs that were ill suited for the high rpm work than 4V allowed. The flexing of the valve train components was often enough to offset the benefits of the larger valve perimeter. As designed, the included valve angles were all over the place, as were the shapes of the inlet tracts with multiple right angle turns from the carb to the combustion chamber. Studying the faults isn't meant to be derogatory. I quite enjoy all the oddball engineering designs. Dead ends are sometimes the funnest roads to ride.

But as a long time Exclesior rider, I'd never describe the bikes as having "excellent steering" as Mr. Taymans did. ;)

Anonymous said...

Thanks, Paul.

What a technically literate discussion, even if I don't understand all of it. On a par if not better than what we see in the best British classic bike magazines. I've got a friend who's got engineering credentials I can't even dream of, so I want to pass this post/discussion along to him.

BTW, happy to see the G45 Matchless cross 50K US at Stafford. Maybe the new bottom line for this underdeveloped, underappreciated AMC racer, still the most gorgeous competition two-wheeler that ever graced a race track.


Jason Cormier said...

There is something to be said about our conceptions of these early attempts, something obviously coloured by our modern perspective. Paul makes a good point that I hadn't considered, but everything I've understood placed the discussion in parameters of "they were on the right track, but the execution wasn't correct because they didn't understand what we know now...".

In that vein, we have this idea (which I have fallen prey to) of the DFV being the turning point where suddenly 4V chambers "made sense" and performed far better than the alternatives, as if no prior design had explored the flat/pent roof chamber and thus never took advantage of squish etc. Odds are the Cosworth gang weren't the first to discover or execute these principles, but they were the ones who succeeded with it and got the glory.

I will stand by my combustion efficiency statement though - this is an area that is often overlooked in discussions of engine design. Squish, axial swirl, spark, anything that contributes to a good burn in the cylinder. The TLDR summary is you need to have turbulent flow combined with the piston squishing the mix toward the source of ignition to get optimal burn. It's a key part of tuning that is often forgotten in the unending pursuit of flow through the head - pumping air is well and good, but you have to make the mix go bang properly to get your best power. Hence trends like sticking extra spark plugs wherever they fit so you can introduce multiple flame fronts and get a more thorough burn - a consideration for cleaner emissions as much as for making power.

The Vintagent said...

Hi Jason,

I've written several articles on combustion theory - ant theorists - including Harry Weslake:

Ricardo laid out the first theories, and Weslake was the first to properly test them, creating his own cylinder head gas-flow test rig in the late 1920s, while working for Sunbeam. Weslake was the first to properly sort out 'swirl', which I lay out in the article, and used what he knew for Norton postwar to improve the Manx, even though Joe Craig detested him!

Anyway, Rudge was the first motorcycle to use a pent-roof combustion chamber with 4 valves, and the principles of 'squish' were laid out by Ricardo, and used extensively in racing motorcycles after Leo Kuzmicki (a professor of internal combustion in Poland pre-war) joined Norton after the war.

Clearly, we're geeks, but I find this stuff fascinating! Kevin Cameron can and has talked circles around me during conversations about combustion, but it's still a fun topic.

Angela Killpack said...

That makes a lot of sense to me to have multiple valves rather than one large one. You cover more area and there is potential for less breakage of the valves. I wouldn't want a valve breaking in my engine if I can help it!