Port Design on Horizontally Opposed Engines 3

I don't know of any production engines that make use of the port configuration you described. I'm sure it's possible, but it would seem to make the job of valve train packaging/actuation more difficult and result in a less than optimal sparkplug placement within the combustion chamber. Not to mention further adding to the packaging difficulties of an already difficult to package motor.

I would be interested if someone could provide a working example of said design.

The answer to Younglearner's question is "Yes".

In the days when engine speeds were lower and most competition engines were only 2-valves per cylinder, the prevailing idea was to fit valves as large as possible for best breathing. This enforced large included angles between the valves, usually in the range of 60 to 100 degrees, which in turn forced the design of separated OHC cam boxes or towers. Once you have large spacing between the camshafts, and only have to fit in a spark plug, there can be just enough room to squeeze in the intake tract. So assuming you have a conventional in-line engine, you can have vertical intake tracts. You also have an almost straight shot at the intake port, with the added benefit that induction is aided by gravity (most beneficial for naturally aspirated engines), and the incoming charge is directed away from the hot exhaust valve thus minimizing density loss from that heat source.

There have been performance oriented engines in the past with this architecture, for instance the pre-war 2.0 Litre BMW engine, and the post-war Bristol derivative. Also, the straight 8-cylinder 2.5 L desmodromic Mercedes engine in the 1954 racing cars had this feature too, and I am sure there must have been others.

However, as Bryan points out above, horizontally opposed engines have different design constraints. For packaging reasons, it is usual to try to minimise the engine width as much as possible, and so such engines are usually short stroke, and accessories are arranged so that the cam covers are the widest dimension. This being the case, there are two things that mitigate against the adoption of "downdraught" intakes of the type discussed above: The first being simply that now the engine width is increased significantly on an engine that is already wide, and secondly with the cylinder now horizontal, the conventional "side" intake now becomes the direct equivalent of your downdraught port with all it's theoretical advantages, and your downdraught architecture is now the less favourable horizontal intake.

All in all, one would think that you would have to be crazy to design a high performance boxer engine with the porting arrangement in your question, unless you had no constraint on engine width, or perhaps if it were pressure charged. But as it happens, there have been racing engines with this design. The perfect example that comes to mind is the 1966 BRM H-16 engine, but in this case the architecture was forced upon them since the upper bank got in the way of the preferred porting for the lower bank. Thus both banks on both sides had the "between-the-cams" porting. A picture of this engine can be found at: - However, an altogether much better review of this engine (and 49 other racing engines from 1913 to 1994) is to be found in the Karl Ludvigsen book "Classic Racing Engines", Haynes Publishing, 2001; ISBN 1 85960 649 0. I highly recommend this book to anyone who is interested in engine design, and particularly historic competition engines.

I think it unlikely that any modern engines would adopt this between-the-cams porting for two reasons: 1) Since the Gurney-Weslake V12 engine of 1967, racing engines have had much narrower valve included angles (typically around 30¡) than in the past, for more compact chambers and better combustion efficiency, so that there is no room for downdraught porting. 2) With port runners and a plenum sitting above the cambox on a conventional in-line engine, it makes for a very tall package.

PJGD

PJGD,

Very good stuff. I was aware of a few engines that used downdraught ports, and you pointed a few new ones out to me. I had been unaware that the design had ever been used on a horizontally opposed engine. All-in-all, good info.

Regards,

Bryan Carter

I think another example was the early to mid 60's Borgward.

I agree it is of no benifit to a boxer, as the standard layout used by Subaru, Porsch and VW gives a downdraft port and excellent torque and throttle response as a concequence, especially when multicarbied.

There is nothing better for throttle response on a carby, NA engine than being able to look down through the wide opened throttle and see half the back of the inlet valve.

PJGD
Thanks for the tip on the book. It is now on my must have list

Regards
pat

Pat:

The Borgward RS 1.5 Litre is one of the engines described in Ludvigsens book. If this is the engine that you are referring to, it has conventional porting. But there are plenty of other neat things about it, for instance it had direct (in-cylinder) fuel injection, and it was an aluminium "through-bolt" engine.

PJGD

I was still in school when I saw one, and the through the top of the head porting is anecdotal from an unreliable source.

I really must get that book, even if only to remind me to chech all detsails from that source, who unfortunately is VERY OPINIONATED, and is a close family member.

I still can't convince him that a SBChev is a 90 deg V8. He is adamant that it is 60 deg, even with a bare block sitting in front of him.

Regards
pat

I know my questions seem very crazy, believe me my handle fits me nicely, but I was fleshing out the basics of a chassis on a $20 Autocad program (I'm military what do you expect?) But the chassis concept is very unorthidox. It took the idea of horizontally opposed engines and cab-over trucks. I figured why not make a firewall/floor and put the engine and drivetrain under it. I have seen many ways to use this to lower the CG. i.e. Multi-disk clutch and dry sump oiling. Anyway the engine compartment ended up being 57.92 inches wide in it's widest point, which happened to be long enough to easily fit an H-12. Any who I thought that perhaps a between-the-cam port configuration could perhaps aid in the engine's "squat" positioning. I have enjoyed all of your responses and your openness to my questions. It's nice not getting dumb looks for a change.
Maybe one day I'll have a teacher to bother .

I too think that there is a lot of merit in a commercial vehicle layout in which a horizontally-opposed engine is arranged in unit with both the transmission and the final drive, and mounted underneath the frame rails of, for instance, a tractor/trailer. The weight distribution might be wrong for a long wheelbase truck. Cardan shafts would then take the drive to the independantly suspended wheels, possibly with hub reduction gears. Would you believe that Bugatti exhibited a concept truck diesel engine along these lines back in the '50's!

But this dreaming is off topic. More to the point is that all trucks, military included, have diesel engines where the combustion chamber designs are such that there is little opportunity for widely spaced camshafts and available space for "vertical" ports, since the valves are themselves vertical. Additionally, all diesels for this duty will be pressure charged and therefore do not need long tuned runners. A simple plenum manifold above the ports, which would not take up much height on your horizontal engine, would work fine.

PJGD

Aircraft piston engines use vertical ports. Sometimes the intake manifold is on top of the engine and the exhaust on the bottom, and sometimes both intake and exhaust are on the bottom. There is less room on top because the pilot must be able to see over the nose, and there is lots of room on the bottom because of the ground clearance needed for the propeller. Depending on the aircraft, the width of a horizontal engine is more easily acommodated than height. There are limits, so aircraft engines are large bore, short stroke.

I have a truck with just such a layout.

It is a 1913 Little Giant.

It has a 2 cyl, 2.5 litre boxer engine, with the engine being mounted by bolting the cylinder heads directly to the chassis rails.

The engine is connected to the transmission by a very short shaft, maybe 6" long.

It has a 2 speed epicyclic transaxle mounted mid chassis with the half shaft outer housings also solidly mounted to the chassis rails.

The drive from the half shafts to the rear wheels is by chains.

It is a cab (or should I say seat) over design, with the engine directly under the seat.

It has exceptionally good stability and handling for it's day.

Is there nothing new ???

Regards
pat

This is a very interesting thread to me as I’ve been designing a modular horizontally opposed 12 / horizontal in-line 6 (similar to BMW K1200) engine using this particular architecture for some time now. The only application in which I’ve seen the combination of a downdraught inlet port and a completely horizontal engine architecture is the Mercedes M291 which debuted in the 1991 Sauber C291 Group C (Le Mans) car. It was a twin-cam 4-valve, normally aspirated unit displacing 3.5L and made roughly 700 bhp. Exhausts, naturally, exited from the top of the engine to facilitate the lowest possible mounting in the chassis, but there are also serious heat rejection benefits from this sort of packaging as well! Shoot me an e-mail (variant@qwest.net) and I’ll send you a rather good, hi-resolution picture of the engine.

As for cons, I was reading somewhere on a Jaguar related tech site that one issue with the downdraught port design is obtaining a low disposition angle between the inlet ports and the valve axis. The Baily designed Jag V12 was able to achieve a ~30 degree disposition between the valve axis and downdraught port (much better than the old M-B SL design), and the resultant volumetric efficiencies were acceptable by racing standards and pretty good by road car standards. That 2-valve engine was fairly square in nature and I don’t think flows would be an issue in a more oversquare, 4-valve design such as the Mercedes or mine… unless the engine could rev extremely high.

As an addendum to that, I was able to bring the valve/port angle down further by employing a radial disposition between corresponding valves, i.e. a geometric splay between corresponding inlet and exhaust valves. This requires a conical profile cut on the cam lobes in order to actuate them. The radial disposition results in lobe couples that are now closer to the bore center midpoints than the bore centers themselves, and I have logically located the cam bearing support webs there. This, in addition to a radial dual-ignition agreement, leaves a conical void directly over the center of the bore to fashion a shallow, straight inlet, with plenty of cross-sectional area. Also as a note, the resultant included angle between inlet and exhaust valves is a approximately 29 degrees, and the valve actuation system benefits from the employment of stud mounted finger followers. I’ve rudimentarily modeled all of the relevant geometry up in CAD to verify that there are no interference/packaging issues and am in the process of completing two dimensional sections, projections, and detail drawings from that. There are still many design features yet to be resolved, but I’m pretty sure that the head and block structures could be finalized and prototyped, given capitol and the proper powertrain engineers.

The idea was to create a large (i.e. powerful) engine package that could be centrally located in lightweight performance car platforms (of which, my associates and I have conceived two thus far). Myself, I fancy the idea of a very lightweight sports car with a large, powerful mill located so as to facilitate the lowest CG and most centralized polar moment possible. A downdraught ported, horizontally oriented powerplant seemed the natural solution. We also contemplated applying the design to marine and truck applications, where, much like sports cars, packaging space is less of a concern. If you’re interested in seeing any drawings as I complete them, please just inquire. We can establish a dialogue via e-mail if you’d like.

Ryan R. Miller

I went back to my data to recheck the info on the Baily Jag engine and found myself in error. That's what I get for working from memory! The downdraught Jaguar V12 was, in fact, a 4-valve design, not a 2-valver. However, I also found that the disposition from inlet tract to valve axis was closer to 40-degrees, not 30 as I had suggested. I apologize for the errors.

It's also worth mentioning (as it slipped my mind earler) that there is at least one manufacturer using a downdraught port configuration in an engine. Mitsubisi's GDI (Gasoline Direct Injection) engines utilize a downdrought port, though the GDI technology is not available here in the United States.

Ryan R. Miller

Ryan;

That is interesting: I had forgotten that the original V12 Jaguar was a downdraught port design, but looking back at the Wally Hassan SAE paper (720163), I see that it was. Not so sure about it being a 4-valve head though. Neither the SAE paper, or Hassan's autobiography (Climax in Coventry) mention 4-valves, and indeed they imply it was a 2-valver. Perhaps they also built a 4-valve prototype to see what it did for them.

Hassan notes that the engine performance was disappointing (only 100 bhp/litre from 5.0 Litres), and seems to think the ports were to blame. He mentions that the Mercedes M196 engine with the downdraught ports also did not live up to it's promise, due he thought, to ports and valves being too large.

So, as I understand your post, your engine is a boxer with radial valves, and the large space in the center that radial valves provide permits the between-the-cams downdraught (but actually horizontal) inlet ports? Have you gone for the exhaust out of the top or the bottom? Are the adjacent ports paired (conventional), or the diametric opposite ports paired in function as on the BMW M10 radial valve engine? I am assuming conventional, not diametric porting, and inlets on top with exhausts exiting below. This way you can get a good straight shot at the inlet valves, and the ports are not really horizontal at all but probably closer to the vertical. But it does mean that the air flow converges on entry to the cylinder which may not give optimum breathing.

The cam profile on an "angle" would bother me. All previous radial valve engines that I have been aware of have managed to avoid that solution to the valvegear geometry problem. As I understand it, your camshafts are out of the way of the ports, and the valves are actuated by rockers?

PJGD