Larry Widmer's Soft Head

[Pinhead]

I've been very interested in Larry Widmer's Soft Head design ever since I read about it a few years ago. I'm going to make this post as a compilation of the information that I've got so far. If anyone has any more info about this tech, please add to this thread. Also note that I'm far from an expert on this, it is just my interpretation; if anyone knows that I'm wrong please point me in the right direction.

Also note that this first post is a copy/paste of some threads that I started before I discovered GFN. Some of the stuff will be somewhat contradictory to what I've learned on here about how real engine builders work, so I'll slowly edit stuff that I now know is wrong out.

This is from a Hot Rod article circa 1985. It's about Larry Widmer's Soft Head design.

Obviously this is a highly complex, integrated system with more nuances than we could possibly explain here. For instance, Larry states that valve lift and timing should be dictated by the shape of the piston dome you want, rather than vice versa, and his motors use very conservative cam profiles. Yet, because of the port design and velocities of his heads, they will flow 95% as much at .450-inch lift as at .800-inch. He also uses his own "super critical area rule" (a computer formula) to reduce exhaust port volume to increase exhaust port flow and low rpm torque.

Does it work? On a 498-cubic-inch AR Ford Hemi test engine Larry ran 16.2:1 compression; a 278/280-degree (at .050), .656-inch lift, 114-degree lobe center cam; and 29 degrees of ignition lead - he calls this a typical engine. The results at 5700 rpm: 860 hp, 790 ft-lbs of torque, exhaust gas temperatures of 809 to 819 degrees F, and a BSFC of .347 (Both the EGT and BSFC are incredibly low; normal would be about 1400?F and .550).

This is my interpretation of the Soft Head technology after rolling it about in my head for the last few weeks. His website says very little about the camshaft, so the camshaft part is how I would do it.

EDIT: It has come to my attention that my interpretation is probably wrong in this matter. Ignore my ramblings in this post and pay attention to the articles.

Many engine builders get hung up on horsepower, breathing, and static compression ratio. Generally many builders start with a CR, say 10.5:1 and build from there. They try to cram as much air into the chamber as possible, and then later try to figure out how to keep the thing from detonating, generally by running more expensive higher octane fuel. While these engines have great volumetric efficiency, and can put out decent horsepower, they generally have terrible thermal efficiency (though they're usually better than their lower-compression twins).

Larry Windmer at Endyn, Inc. (The Old One - Energy Dynamics) takes a completely different approach. He builds his engines with as high of static compression ratio as the components will allow. These ratios are up in the 23:1 for low-octane gasoline engines. This is done to make the combustion chamber as small as possible. When the air/fuel is in such a small space, it takes very little time for the flame front to travel from the spark plug to the walls of the chamber and consume all of the available fuel. He also runs these engines in "lean burn" mode. The air/fuel ratios are in the 18:1 range. Since the chamber is so small, the slower-burning lean mixture still burns much faster than a "perfect" 14.7:1 ratio does in a conventional chamber. How is this high CR kept from detonating?? It's all in the camshaft.

I keep reading about people saying boost this, port that, get that chamber full of air. While this can work, it has its limits. With Larry's design, the engine produces more horsepower while consuming less air and fuel. It's a MUCH more thermally efficient engine.

The camshaft is set up like this:

The intake valve starts opening right at TDC of the intake stroke, following the piston down into the chamber. It differs, however, because the intake valve is actually closed before bottom dead center. This is the key. The valve is closed at a time as to put the "dynamic compression ratio" at a tolerable level. However, since there's such a small chamber and therefore a fast burn and very little ignition advance, this compression ratio can still be relatively high, in the 11:1 range or higher without detonation on low-octane fuel (depending on the chamber shape, turbulence, swirl, etc.).

The exhaust valve is opened at about 10 degrees BBDC on the exhaust stroke and closed as late as possible, following the piston up during the last few degrees of the exhaust stroke. The goal is to erradicate as much exhaust gasses from the chamber as possible. 10 deg BBDC is chosen because around 14 degrees BBDC, very little pressure on the piston is actually changed to kinetic energy on the crankshaft (the movement of the piston is a sine wave).

Setting up an engine in this manner has a number of technical advantages over a "conventional" 4-stroke engine. The first has been already mentioned: Small chamber size. This allows for a very fast, controlled burn and very little ignition advance, which greatly reduces pumping losses and thermal losses into the engine block, as well as virtually eliminating detonation. The reduction in pumping losses greatly increases low-RPM torque. It also allows the engine to run much higher in the RPM range.

The next advantage has to do with the timing of the valves. Since the intake valve closes before bottom dead center, the effective intake stroke is shorter than the exhaust stroke. This allows the air to expand further than it was compressed. This is advantageous because hot air naturally wants to take up more space than the same mass of cool air. The further the air is allowed to expand, the more power is extracted from that air and the less heat is lost in the exhaust and engine block. This translates to lower EGT's and longer-living exhaust valves.

From a discussion at MPGResearch:

Originally Posted by automotivebreath

I have been reading about a piston designed by Larry Widmer for the
VW engine called "Super Squish Piston". People are running 13:1
compression on pump gas with air cooled engines, impressive. The people
that sell the pistons require a non-disclosure agreement signature so
details are sketchy. A piston with a dome on the intake valve side and
altered camshaft timing are parts of the equation.

Here's a link to discussion:

http://www.cal-look.com/index2.html

And a picture of the piston from Speed Talk.

Originally Posted by automotivebreath

I will be using Larry's soft head concept on some of my engines. I dont
completely understand what he does so I'm always searching for clues.
Looking at some of the things he has done, it becomes apparent that the
piston top plays a huge role. In most instances it looks as though that he
uses the dome to ether create mixture motion or direct the squish action
into a specific area or perhaps a combination of both. His accomplishments
are extraordinary and appear to be something that can be reproduced.


In the first picture from the soft head article, the piston squish areas are
very defined and I believe conform closely to the shape of the cylinder
head. Squish action is not allowed to travel linearly across the piston top.

In the second example the dome obviously forces mixture movement out
of the area beneath the intake valve, thus creating an effect similar to the
May head mentioned earlier.

That's all I've got for now, will likely add some info in the future.

[Brian P]

Nitpick, your description of the cam events isn't consistent with his specification of a 280 degree camshaft with 114 degree lobe centers (that would imply intake valve opening 26 degrees BTDC and closing 74 degrees ABDC). BUT ...

A similar effect (reduction of effective compression ratio) works if you leave the intake valve open long after BDC than shutting it before BDC, but with easier-to-achieve cam profiles and valvetrain dynamics. Should note at this point that this is commonly called the Atkinson cycle, and the Toyota Prius engine uses it (along with a higher-than-normal mechanical compression ratio).

I understand a lot of what Larry is talking about and the purposes behind them, but still I can't help but wonder why this approach (particularly the combustion chamber design) hasn't gone into production engines. We sure need the best BSFC's that we can get right now. Something tells me that there is a hidden down-side to this approach that makes it a deal breaker.

[302tt]

Originally Posted by Brian P

A similar effect (reduction of effective compression ratio) works if you leave the intake valve open long after BDC than shutting it before BDC, but with easier-to-achieve cam profiles and valvetrain dynamics. Should note at this point that this is commonly called the Atkinson cycle, and the Toyota Prius engine uses it (along with a higher-than-normal mechanical compression ratio).

The Atkinson cycle is good for BSFC but not BMEP. Why would you want to use that on a racing engine? I think there must be more to the technique than that.

[automotivebreath]

Originally Posted by Brian P

... We sure need the best BSFC's that we can get right now. Something tells me that
there is a hidden down-side to this approach that makes it a deal breaker.

Modern productions engines rely on unburned fuel to fire the CAT. To me this means
the engineers design engines where combustion is incomplete (less than 100% mass
fraction burned before the exahust valve opens). Perhaps what's best to meet current
emission's standards is not best for ideal BSFC?

If this is fact then the goal would be to archive complete combustion (at a favorable
crank angle) with reduced emissions.

[Pinhead]

Originally Posted by Brian P

Nitpick, your description of the cam events isn't consistent with his specification of a 280 degree camshaft with 114 degree lobe centers (that would imply intake valve opening 26 degrees BTDC and closing 74 degrees ABDC). BUT ...

As I said, I don't fully comprehend every aspect of the design. Cam timing is definitely not my forte.

Originally Posted by Brian P

I understand a lot of what Larry is talking about and the purposes behind them, but still I can't help but wonder why this approach (particularly the combustion chamber design) hasn't gone into production engines. We sure need the best BSFC's that we can get right now. Something tells me that there is a hidden down-side to this approach that makes it a deal breaker.

It's probably the evil Catalytic Converter's fault.

[Brian P]

302tt, it's quite true that an Atkinson cycle is not good for BMEP, but an interesting consequence to the late-intake-valve-closure arrangement is that (as long as the intake duct is tuned properly) the volumetric efficiency and therefore the effective compression increase as revs increase. So, you get a lower effective compression ratio under the conditions when detonation is at its worst (low revs). The Prius engine doesn't exactly take advantage of this, but Larry might have ...

Regarding catalyst efficiency, it's possible that this has something to do with it but I'm not so sure. A diesel engine is capable of lighting-up an oxidation catalyst, and those have lower EGT's at part load than gasoline engines. If you need to light-up a catalyst, there are creative ways to do it that would still allow the high compression to be utilized.

Random thoughts. Big squish bands done properly are good for burning the main part of the mixture through turbulence but they also generally mean a lot of quench (think high HC emissions) and the last bit of mixture in the far reaches of the squish/quench area can detonate just as much as anything else. High squish turbulence usually means high scrubbing-action and therefore heat transfer (heat losses).

There has to be something else to it.

[automotivebreath]

Bryan,
I'm certain there's more to it.

One possability; increase flame intensity (create a hotter flame) to a point where
combustion is complete by ~ 15 - 20 degrees ATDC. HC emissions would be low, but
what about NOX? Even with "creative" ways of firing the CAT (read expensive), would
the CAT handle the high levels of NOx? Isn't that why the 45 MPG TDI diesels were
banned in the US?

[rrrobert]

ok let me get this right ...
1) no independant tests of said engine design .
2) nobody apart from the designer understands it ,
3)and he cant explain it in a way that can be understood..
4) it makes no sense
5) emperors new clothes ?

[automotivebreath]

Hi Robert,
I feel I understand the concept fairly well, although I'm sure there is more than meets
the eye. To me it makes good sense, traditional in cylinder geometries provide
inadequate combustion.

Perhaps he can explain it in a way that others would understand, but has no desire to.
Creating what appears to be a very close match of the bottom of the cylinder head
shape on the piston top looks to be very time consuming!

[GIGAPUNK]

This goes beyond vague. It's down right incorrect. In the first post it's explained that the high static compression is made possible by the dynamic compression that results from closing the intake valve right near BDC. That would result in HIGHER not lower dynamic compresion! My 427 has an IVC @ 48 ABDC. This gives me a dynamic comp of 8.9 from my static 11.7, if I were to have a IVC @ BDC instead my dynamic comp would equal my static comp. ie. my dynamic comp ratio would INCREASE from 8.9 to 11.7. It's the compression that's not happening with IVCs ABDC that lowers your dynamic comp. As describedin the first post... this makes no sense.