"Thump-8" 4 stroke 1/8th scale engine
 

I figured I would make a thread for my new engine prototype instead of cluttering up the FS26 thread, so here goes!

I named the thread (and for now, the engine) the "Thump-8" because I'm not sure what the final displacement will be. The initial prototype is looking like it'll be about .39ci, but that number can easily change with slight variances in bore and stroke without changing the overall design, depending on how much power and RPM I can wring out of it.

The goal of this design is to end up with a 4 stroke engine that bolts directly into any place a current .21 two-stroke fits, and is able to use the same (or at least very similar) gearing and clutch hardware. The reality may be slightly different than that, but that's the point - to find out if it's possible. Nobody has ever attempted this before (that I know of). It will need to be compact, VERY high-revving and if I can manage, I would like it to not require a cooling fan... Pie-in-the-sky stuff for sure, but the project should be at least lots of fun.:D

I haven't made metal chips fly yet, but I have discovered Google SketchUp, which allows me to model my engine in 3 dimensions. It's pretty easy to use, and I've already modeled a good portion of the valve system. As I continue to learn how to use the program, I can see it's going to benefit the design and engineering process in a big way. Much better than drawing in 2 dimensions and just trying to imagine how it would work in 3-D. :cool:

Tonight, I get to pick up my prototype piston, ring, wristpin, and sleeve. Once that stuff gets digitized, I'll have something concrete to base some dimensions off of.

any clues to the stroke ? 16.4mm from the 26or smaller ?

Wasn't the now defunct FS-40SC supposed to be the "1/8th" scale 2 stroke subsitute?

Yeah, in the same way that a bottom of the line Honda with a Type R sticker slapped on it would make a good substitute for a race car. While the "R" on the sticker seems to indicate it would be competitive in a race class like the GT2 class in LeMans, it really can't because it was never meant to.

My theory is that most companies that already sell 2 stroke RC car engines have no reason to make competing 4 stroke engines. Here's a hypothetical: Let's say one of the popular companies makes a 4 stroke that is actually competitive with their 2 strokes. They now have to make 2 varieties of engine, parts support, keep more inventory on hand, and it takes sales away from their already successful 2 stroke engine lines. They would likely lose money, especially once another company decides to compete, prices drop, and sales are spread thinner among the competing companies and engines.

This is why I think the 4 stroke "car" engines we've seen are all slightly modified airplane engines that fit squarely into the novelty market, not competing directly with the mainstream RC car engine market. Yes, with enough work, you can make them competitive - but it was never their intended purpose so it costs more than it should and it's not easy by any stretch of the imagination.

The only companies it could possibly make sense for would be companies which don't currently sell to the RC car market, such as Saito or YS... but even they would be taking a big chance trying to create a market for an entirely new (and expensive) product. If they wanted to do that, I think they would have by now.

I'm doing this for a completely different reason. I'm not motivated by money. I simply want what I want - it's my passion, my hobby. I want to do it right from the very start. I'm not starting with an existing engine. I started with a blank sheet of paper, some design goals, and a big cup of coffee. I don't have a corporation to answer to. No deadlines. No investors. No politics. Most importantly, no risks. It's a hobby for me, so if it doesn't work... Eh, no big deal. I'm sure I would have found another way to spend that money, and it may not have been as fun or satisfying.

I would like to see the outcome as well. Good luck on your endeavor.

The sales of engines in r/c car market is driven very much by the horsepower ratings published by the manufacturer . I know when I started out eons ago, the I remember asking the store owner, "Gimmie the fastest, most powerful engine you got!". And I'm pretty much sure that's how many still buy their engines.

By their nature 4 strokes can never bost the same hp figures as the 2 strokes, in addition 2 strokes are cheaper to build, have fewer moving parts, need less parts support etc.

I guess it boils down to this.... r/c car and boat people want hp/rpms which = 2 stroke screamer. Plane people need (notice I said 'need' vs' 'want) torque = thumpers.

Personally, I'm now way past sky high hp figures & I go for realism/novelty. I don't race competitively so this is usually not a problem. Right now, I'm at the stage that if I can squeeze a thumper in place of a 2 stroke, I'd do it!

So DP, if you do come up with a new reliable 4 stroke, I'd certainly be willing to buy one from you.

You know the term "floating valves" ?
With surch high rpm's a very strong return spring is needed to close the valves folowing the camshaft. Having such a pressure and so resistance on the camshaft you will loose power.....

The only thing to solve this is using rotating valves. Once I have seen an engine using 2 rotating tubes above the cylinder, one with an hole to the cylinder to let the fuel through the tube in and one tube with a hole conected to the exhaust.

Yup. I've seen all the aircraft, motorcycle, and automotive non-poppet valve designs (that I can find online), and all the RC non-poppet valve 4 stroke designs. What I'm drawing is much different than anything I've seen before. The only reciprocating thing in my engine will be the piston. :D

I've had this design kicking around in my brain for years but I assumed it wouldn't work, or that I would never have the means to make one. Now that I'm seeing it in 3 dimensions, I know it will very likely work and that it isn't outside my abilities and means to at least make some prototypes.

we've been over this it's pneumatic valves springs for high rpms.
a metal spring horribly slow compared to pneumatics.
drag is usually reduced by rollers on the cam, rocker or, lifters
a big tube with holes lol that sounds so silly but yet probably the easiest to seal the valves
you'd till have the drag of a chain, belt or gear, in that.

just make sure we can use turbo plugs on this compatibility is the key Dp-buggy wants here

HMMM! I hadn't thought of that before. I wonder if there's such a thing as a 4 stroke turbo plug. I have been drawing everything with an O.S. Type F plug so far.

Does anyone know what the other 4 stroke manufacturers use, and if they have anything that is equivalent to a turbo plug?

regular plugs with no idle bar. the turbo plug is 5/8" while a normal plug is 1/4" in diameter.

A rotary disk valve or tube valve will certainly work and best of all with our fuel there's no problem with exhaust valve's cooling and lubrication, but if you want ultimate power you need a cam, only in a cam profile you can obtain really high area under the curve - aka WORK, since it gives a non-linear relation between valve position and crankshaft. With a rotary type valva you can only go to a certain point dictated by the linearity of a 360deg circle - the crankshaft. That's the reason you dont see the F1 guys messing with rotary valves, and maybe thus even 2-stroke engines it would actually rob them a ton of power.

X

RC 2 stroke engines do use rotary valves for their intake. That's what the hole in the crankshaft directly under the carburetor is.

The design I am developing has lots of area under the curve, and lots of time with the valve ports fully open. I need to plot the curve and compare it to a typical engine's curve to see the real difference, but I think it's going to be pretty dramatic. It also has no transitional time to get the valve moving and bring it to a stop more gradually like a poppet valve engine has to have. If a poppet valve just slammed open and shut, it (and its valve seat) wouldn't last long.

In other words, the bottom and top of the curve in the disk valve engine has sharp angles, and the "ramp" is a constant angle, and it has a substantial flat top too.

Now I really want to graph it to see what it looks like, given the valve timing I've chosen. I'll post it here so you guys can see what I'm talking about.

The one thing that's going to be hard to compare will be the cross-sectional area of the valve opening. It'll be easy to figure out on the rotary valve, but I'm not sure how to calculate it for a poppet valve.... unless you just calculate the area of the outer surface of the cylinder created by the movement of the outside of the valve face, using the valve face diameter as the cylinder diameter...

I'm not thinking the poppet valve stands a chance once the cross-sectional area is taken into account.:sneaky:

LOL, what I ment is that maybe thats one of the reasons F1 doesn't mess with 2 strokes either.

That's what I'm talking about, in 2 stroke rotary valve engines like ours for example, the gains that you get from the rotary valve design against a piston port for example are inmense, the same goes from a rotary valve to a cam curve, they are usually worlds apart, I'm not saying you are going to find a "brick wall on that road", im saying i've walked that road and found one! I hope you find a way to go over it though :nod:

The thing is that a poppet valve is not moved by the circular motion of the cam's shaft, instead is actuated by the cam, and that allows for independence in the design of the curve, allowing you to plot a curve that is non linear to the camshaft and crankshaft rotation, unlike any rotary valve design on 2 strokes or tube or rotary disk valves in 4 strokes

And thats kind of what I thought, but I even tryed very aggressive engines like hot on-road .21's, the area under the curve is not that hot at all compared to a hot 4 stroke, and if you try to get more agressive you start messing up and mixing the cycle stages.

Also the same thing allows for not needing to have that big of a port/valve area because the valve is fully open more time, and that helps velocity and torque.

Just for help: 70 to 75% of the poppet valve's face area is a good and conservative aproximation ;), that's what (+/-) a good full scale racing engine flows through a puppet valve like a modern (ok not new LOL) nascar or pro-stock engine. So there you go: valve area x .725

good luck, I would love to see your racing 4stroke project :sneaky:

X

Cool! I didn't do much last night - hopefully I'll be more motivated tonight.

I thought about the need to maintain port velocity too. I'll likely end up changing the shape of the ports depending on how it works. The good thing about the prototype design I have is I can easily change the port shape, size, overlap, and timing by swapping out the upper valve seat (which is basically just a flat piece of steel with holes in it).

I even had thoughts of a guillotine-style power valve arrangement to vary the size of the ports depending on RPM. It wouldn't actually be all that hard to do, BUT it probably won't happen since the motto of all good RC engines is KISS... I guess I'll file that one under the "Anything is possible" category for later on.

Is that 3D program good enough?

I've worked with Acad, will I need training or is it straight forward enough?

Spherical valve 4 stroke engines

http://www.coatesengine.com/technology.html

forced induction anyone lol.

Thats exactly what I'm talking about.

When an all-out racing engine is what you have in mind a cam will give you more.

Just look at the angle that the air way has in that rotary valve and only that gives back a chance to the poppet valve, and at the end you see the gears moving the shafts that rotate the rotary valves, there's the linearity that will kill it when you put it again side by side with a poppet valve race engine.

They say +14,000rpm, ok, but a hot bike will also do that all day, and a racing 1.6L-2.0L Honda is pretty close too, not to mention F1! and our little 4 strokes!, and way more importantly a poppet valve engine will probably have more power up there because of the timming available with the design. Whats good to be able to spin 20,000rpms if your torque curve is scrapping the floor?

X

Umm dude do some reading on the Coates design, it absolutely smokes a poppet valve engine in performance....not just beats it, smokes it....they did several v-8 engines and the results were absolutely stellar....


heres a test by popular hotrodding on a Coates conversion on a Ford 302

Where the CSRV really shines is in its airflow potential compared to a poppet valve Bench-marking a 5.0 L engine from a Lincoln, the stock Ford casting (when tested at 28 inches of H2O) flowed approximately 180 cfm on the intake port at static. The rotary valve for the engine in comparison flowed a whopping 319-cfm at the same test pressure. Equipped with the poppet valve head, the Lincoln engine dynoed at 260 hp and 249 lb.-ft of torque. When equipped with the CSRV head at the same 5,500 rpm test protocol, it made 475 hp and 454 lb.-ft of torque, with no changes to the block or rotating assembly: The higher power was a result of diminished frictional and pumping losses, but the inherent airflow benefit of the spherical valve was the major contributor. With a conventional poppet valve, it can take 34 degrees of crankshaft rotation or more to reach a fully open position, wasting energy and limiting volumetric efficiency. With the CSRV, a comparable port area is exposed in only 2 degrees of crank rotation. The CSRV allows for superior surface flow coefficients from its spherical shape. With the standard 4-inch Ford bore, the factory poppet valve covers only 15.8 percent of the total bore area, while the rotary valve is measured at 20.5 percent.

Those Coates cylinder heads look so trick! I've always liked the way that design looks. They must have some VERY special materials to make the seal work properly. OK, so now I have the motivation to do some math to find out what kind of numbers my design allows.

So, I ran some numbers and did some graphs. Compared to a .40 size engine with 9.5mm diameter poppet valves, my ports should flow better, just based on cross-sectional area. Not MUCH better though, which made me think of ways to revise my design to get even bigger ports. I'll have to draw my new idea so I can make sure it's physically possible, then run some more numbers and graph it to make sure it did what I wanted.

I'm still pretty sure my original idea would work well, since there's no obstruction in the path of the gas flow. It's like dumping water into an open hole vs. one with a valve face sitting in the middle of it.

I haven't even cut metal yet, and this is already FUN!:D

I am quite sure they saw big gains because they are comparing it with a stock motor. Conventional valved Ford 5.0's have been built that make even more power then that.

It is very easy to make something that will improve a stock Ford's performance. When they bolt it on an F1 motor and start winning races, then I will believe. Yes, poppet valves are old tech and have some major drawbacks but they are still in use for very good reasons. No one has as yet figured out a way to make any sort of rotary valve with a non linear type motion. Until they do, poppet will rule.

I see the best bet for future 4 stroke innovation being electro-mechanical valves. Keep the poppets and their motions but ditch the cam shaft. :D

Something like this?? http://home.cogeco.ca/~davebowesevic/

Yes

EXACTLLY!!!

Common maximus :confused: aples to aples dude! reading...:o? a stock Ford 5.0 flows nothing, but there are enough aftermaket heads for that engine that flow close to 400 cfm. A ported modern Hemi head flows around that too. And max flow is just one part of the equation! the speed that the poppet valve design allows the valve to open, and after it stays fully opened for a while, close, will never be attained with - at least - that design, it's just phisically imposible. It's area under the curve what gives you power. That head flows 319cfm for a degree or two, and is the other 90 deg or whatever following a linear curve from 0cfm to 319 and back to 0, not the case at all for apoppet valve, you can use the first 10~15 deg to raise it to the max for full flow, be there for ~60 deg and bring it down to 0 in the last 15 again and sudenlly the head flowed an average that is huge compared to the rotax head.

LOL, almost all modern stock heads do about 315cfm with good porting.

that type of head can ONLY do maximun flow at precisely the cam position that the valve is fully opened, lets give it 5 degrees for example if you make the hole bigger than the port, a poppet valve can be opened at maximum for whatever degrees you want - it not linear!!! And you cant do the holetoo big because it takes degrees to do it because of its linearity, not the casewith poppet valves.

AND again, I'm not saying is imposible, I'm just saying that as long as the valve opens with a linear relation to the crankshaft/camshaft position, the poppet valve will have a chance to shine if done correctly.

But as the wingman said, the big dogs would be using it if it where that good, at least when full blown racing is what you want.

X

Wow. You are The Brave Man if you try to design RC device called engine in google scketchup. I am not talking you out of it, I just try to immaging how many hours you will spend to design main body of engine, shafts, valves, piston and other small things for motor. I am pretty sure you will be able to check for interfiriences, gas flow, perform tolerances patrol in working temperature conditions, flex of components in working temperature conditions ect.
Anyway-good luck on your project.
I suggest you to contact to MXwrench here in forum, he is moderator for this section-he knows 4 strock RC engines as anybody else and I know he preformed a lot of development on them

With a conventional poppet valve, it can take 34 degrees of crankshaft rotation or more to reach a fully open position, wasting energy and limiting volumetric efficiency. With the CSRV, a comparable port area is exposed in only 2 degrees of crank rotation. The CSRV allows for superior surface flow coefficients from its spherical shape

remeber they made that HP on a bone stock Ford 5.0 from a lincoln.. this is not a race built engine at all, no headers, no high compresion pistons, just a stock motor with this intake technology....Good luck reaching that level of power bolting on just a set of aftermarket heads onto a stock 5.0..

I am unsure where this technology is headed, Coates holds all the patents on it and likely wants some huge money to allow any of the big guns to use it..... I myself wouldn't so quick to write it off, as you have to remeber the testing was done with first prototype heads, not with heads that were developed for racing.... being able to almost double the engines power output and nearly triple its RPM range is nothing to sneeze at..especially considering they are a first generation test head...with proper development this system would by far exceed what they currently are doing...You shouldn't be so quick to write off new technology, as you really don't know anything about it, its limitations or its advantages..... As i say your not going to bolt on a set of aftermerket heads on a bone stock Ford 5.0 from a Lincoln and hit that kind of power, sorry but its not going to happen......I know you can build an engine to hit that power, but this was done on a fully stock Lincoln engine, full emissions, stock pistons, stock exhaust, everything the way it comes from Ford....that is nearly a 100% increase in power and nearly triple the effective RPM ranage.... find me a set of aftermerket heads that can do that please..... then keep in mind this was done on a first generation head LOL you guys seriously crack me up !

The Coates stuff is not new technology. Its at least 15 years old, the article is 10 years old. If it was any good, people would be using it. Money is not an issue with big time racing. The only concern in that arena is HP/TQ advantages. And for the money it would cost for those Coates heads, you could build a full on turbo SVO block small block making upwards of 1500hp.

maximo, I really dont know like you said, I dont have THAT system in front of me, but really, how is it possible that the rotary valve in this design opens completely or enogh for any substantial flow with 2 deg of crank rotaion? look at the picture! just rotate it in your mind with out david blaine in the room and you will see, who cares what they say, its simple mechanics that can be seen from the picture, its almost exactly like opening our 2stroke crankshafts(?), but worst, divided by 2, how can 1 degree of camshaft rotation open up a rotary valve that is about 90 degrees of its circumference? I dont see it man. It will probably take about 90 deg of rotation to open the valve to full flow,then the valve will be there for an instant (ok, unless the valve is expossed completely in the chamber like it is in the upper section of the picture for, I dont know, about 90 deg? - but then compression goes to the trash and piston weight tothe roof trying to get the compression back) and then it will be the other 90 deg closing it.
Only with that disadvantage the system probably can be easely bettered by a good poppet cam, a good roller cam can open a valve at least ~90% of full flow in ~10 -15 deg not 34 like an engine designed for swirl and low end torque. One of us really does not understand something here because to me this is more than simple when applyed, dont get me wrong, it can be me :sweat: its just that I gave a lot of thought to that type of design years ago with big-name people that really know their stuff and at the end of the day there where two things that the system did not do with any effyciency - opening/closing the valves (extremely slow and again linear and cooling /lubricating the exhaust valve, look at the struggles that the wankle engine has had through its life span to keep its act clean and fuel efficient.
In our earlyer designs our rotary valves even had that same exact round shape in the opening and closing faces and then we noticed that the port opens "quicker" only because it robs it from the area under the curve, the thing that gives you all the power. :rolleyes: Im telling you Ive been there.
Im telling you, if it where good the big dogs would be using it, a lot of people do a lot of mind burnouts over that design and end up with the classic power maker. I'm a proud one :D, dont worry, theres space for you :smile:

And just for the record, that head probably requires a different ehaust manifold/header, and different intake manifold, and it obviouslly has a diferent cam, if you put a good head/cam/intake/hearder combination to a 5L ford you can obtain 500 to 550hp, and thats done every day just take alook at any mustang mag. if you do the really trick stuff you can get 750 from 302 cubic inches, again, its been done every day, no turbos or alcohol or nitro, just high octain gasoline, compression and flow.

X

Those are all valid points. However, one can't compare F1 technology to a hobby glow engine. I've yet to see a roller follower in a model engine. Few valve trains for model engines have been optimized as they currently exist. There has been no demand for high RPM four stroke engines for cars. All of the engines which exist are based on the needs of model airplanes, power at low RPM <12kRPM. The OS S-C use only stronger springs and a new cam. The 40 additionally has slightly oversize valves. I don't think it's practical to grind a model sized cam with the kind of tolerances required for the aggressive opening lift curves being thrown around here. Pneumatic valve springs in models? The best we can hope for is forced induction like the patented YS engine.

The only way to improve the opening speed of the rotary valve is to increase it's diameter or axial length, which brings to light other concerns. Disc valves leave few options due to size constraints.

Two stroke model engines are virtually the same for cars and airplanes at the top end of the power/displacement heap.

Coates holds all the patents on this sytem, the big dogs don't have the rights to it, and by what it seems Coates isn't playing ball with them.... I do see how in theory you don't see the system working, however I am smart enough to know that unless I see it in its full detail its not even worth speculating over, good or bad..... There may be more to the system then what we are seeing, the opening and closing are done by a floating valve of some sort, so as I say untill we can see it in full I can't say.....I am not a naysayer type of person, and in fact I find people like you comical ...remember they once thought the world was flat....In fast every great invention once had a critic saying it would not work....... All I can say is there are alot of much smarter people then me, and I know enough to not to discredit someone elses idea till I can see it disproved with my own eyes...all the armchair engineering in the word means nothing if key elements of the design have been kept hidden...So speculate all you want on why it wont work, as one day you may be eating your own words over it .....

By their own website, it is a spherical rotary valve driven by chain, belt, or gears off of the crankshaft. Such a design, regardless of the details will have certain advantages and certain disadvantages to a conventional valve system. This is engineering FACT, not speculation. Now whether or not they have managed to get the advantages to outweigh the disadvantages and result in a truly superior system is unknown. My guess is that they have not but I would be very happy to be wrong.

I am not skeptical because I fear new technology, I love new stuff. But you have to understand that people have been trying to come up with better valve systems for 4 stroke motors for over a hundred years (including similar rotary systems) and ALL have failed to replace poppets in high end performance. This system may very well wind up being great for street cars or sport racing but that doesn't mean it will win at the highest levels. Everyone thought AAC was the next big thing but good old ABC just won the worlds.

Personally, if I was an engineer working on all out high tech race motor design, I think I would be working on some sort of slide valve with electro-mechanical actuation. I certainly wouldn't be wasting my time on any form of crank driven rotary valves. Even if someone gets them to work, mine has more potential on a strictly theoretical standpoint.

One thing to keep in mind when debating all types of rotary valves is this:
Everyone assumes the valves turn at a constant rate. There is a very easy way to vary the speed of the valve, making it open and close MUCH faster, yet slow down when it's closed.

You can speculate all you want, and even when you can plainly see the valve mechanism, you may not know exactly how it operates.

Gear driven rotary parts can turn slow for half-a revolution and fast for the other half when the gears are offset (not centered) on their shafts. :sneaky:

I honestly don't know enough to even begin speculating on what they can or can't do ...However they did show some good results on a stock engine...whether they can be used on a top level race engine is something I have no clue about.... Also without them showing the dyno curve its impossible to know what kind of power curve they achieved...their peak numbers sound impressive, but who knows the full story without a dyno curve..... Apparently another type of rotatry valve was being designed but its use was banned by formula 1....So even if they do find a new technology there is a strong proponent to protect the status qua, so even if something beter is out there we may never see it.... top level racing has been known to restrict technology to keep the playing field equal....So wether top level racing uses a certain technology or not really is no measure of whether the technology is good or not....... I am by no means an expert on the subject, I just found the Coates system pretty cool is all...I just didn't like all the armchair engineers being soo skeptical of something they really haven't analyzed first hand..... There is so much information missing from the website and tests, such as flowbench results and dyno curves that really we know very little about the system one way or the other...Hell the system may be great for making extremely economical engines, who knows....I certainly do not, and chances are the only people who have all the answers are the people from Coates themselves.....

That is true but still there is a problem. That would result in slow movement for around 180 degrees and fast for the other 180 degrees (not exactly, but just for speculation we will go with it). But this isn't much help. What's needed isn't slow when open, fast when closed but rather to have it open very quickly from the point of first opening until it is open a fair amount (let's say about the first 30 degrees of opening) and slow way down for the next 40 degrees or so to keep the port nice and big for a long time and then speed up again for the last 30 degrees or so until closed. Two off center gears will not accomplish this, though I could see SOME gains being made in this way.

Now, I'm not an expert on off center gears so I will admit that it very well may be possible to design a complex set of multiple off center gears to achieve the desired result. It is making my head hurt just thinking about how to do it right but there are smarter people than me out there that might be able to figure out. I sure would like to see such a system. It would definitely be an engineering masterpiece whether it worked well or not.

Is everyone assuming the shape of the rotary valve is a constant symmetrical shape?
What if it wasn't symmetrical? I.E. the valve would have a tear drop shaped opening (like a camshaft lobe) vs a square shaped opening.
Instead of the camshaft supplying the non linear opening and closing of the valve, the valve itself could supply the non linear opening and closing.
I'm no engineer, but the idea just popped into my head.

Actually, I think you would want more of an hourglass type shape. Sort of a double teardrop. You would want the leading edge of the port to be very wide so it will flow well even though most of the port is still closed. Then get more narrow so the port doesn't get huge and kill your velocity. Once the leading edge starts to close you would then want the trailing edge to start getting wider to make up for the area being lost at the leading edge.

What have you put in there, a geneva mechanism?

Damnit, it wont post my diagram right.

Let me put the idea into an RC perspective. Imagine looking down the carb opening on one of our motors to see the intake port on the crankshaft (which is a rotary valve). Instead of the normal rectangular shape you would see something more like this:

BACKPLATE
l l
l l
l l
l"-. .-"l
l . l
l l <------Hourglass shaped crank intake port
l . l
l_-" "-_l
c
r
a
n
k


The overall length of the port would be much longer than normal but you wouldn't want a port that big at full open (it would reduce the velocity of the air/fuel mix killing bottom end torque) so it narrows down as it opens and then widens out as it closes.

I like this idea, I wonder how long before someone steals it :D