# Wankel compression ratio

32 replies to this topic

### #1 gruntguru

gruntguru
• Member

• 5,992 posts
• Joined: January 09

Posted 22 March 2012 - 23:00

WANKEL ROTARY
Do you know that the (geometrical) compression ratio of the Wankel is limited?
Can you make a Wankel having 20:1 compression ratio?
What is the limit?

OK I will bite.
Yes I think there is a geometric limitation to CR in a Wankel. One wall of the chamber (there are 4 and there is no need to consider the two end-plates) needs to be an epitrochoid for the apex seals to follow. The mating wall is the rotor face between the apex seals. To achieve an infinite compression ratio the rotor face would need to match the housing (epitrochoidal) face exactly at TDC. My guess is this shape rotor would clash with the housing at some other point in the cycle.

No idea what the maximum CR might be. The maths is beyond me and I will leave it to someone else to Google or otherwise research the answer.

No doubt the maximum CR would vary with lobe count and I would assume that a Wankel with a lobe count other than three would win the "highest CR" contest.

### #2 PJGD

PJGD
• Member

• 84 posts
• Joined: April 04

Posted 22 March 2012 - 23:07

It will also depend if you are constraining yourself to a single stage of compression or prepared to consider multi-stage compression as in the 2-stage Rolls-Royce "Cottage-loaf" Wankel compression ignition engine.

PJGD

### #3 Fondles

Fondles
• Member

• 52 posts
• Joined: July 11

Posted 23 March 2012 - 04:03

OK I will bite.
Yes I think there is a geometric limitation to CR in a Wankel. One wall of the chamber (there are 4 and there is no need to consider the two end-plates) needs to be an epitrochoid for the apex seals to follow. The mating wall is the rotor face between the apex seals. To achieve an infinite compression ratio the rotor face would need to match the housing (epitrochoidal) face exactly at TDC. My guess is this shape rotor would clash with the housing at some other point in the cycle.

No idea what the maximum CR might be. The maths is beyond me and I will leave it to someone else to Google or otherwise research the answer.

No doubt the maximum CR would vary with lobe count and I would assume that a Wankel with a lobe count other than three would win the "highest CR" contest.

The rotors have shallow channels of sorts in them, if you were to make those even more shallow the CR would jump up a fair bit. I don't know enough about rotaries to know how that would affect the flame path in the moving combustion chamber though and suspect it may not be a good thing.
I do know they run pretty high exhaust gas temperatures though, and people that have increased the CR made the EGT jump up even higher. The higher temps were (apparently) not so good for the engine.

I'd like to see one run on E85 though, and with a bit of boost. Should make for interesting numbers.

### #4 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 23 March 2012 - 05:10

The bulk of the combustion chamber is in the rotor and they have different shapes and different cc's similar to the mods for a piston engine.

I note the later ones have a side port exhaust port just like the inlet rather than the bulk of them over the years with a "straight hole" to delay and lengthen the path of the exhaust so the apex seals can do more damage on the way out.

Here's one of the better animations ..

In answer to Man's question, 20: 1 is achievable on a common Mazda Rotary and agree with Grunt that it's down to shape ultimately.. however with the rather large heat path upping the compression too far on pump petrol (not "gas" because gas is .... ) is just a waste of time as the heat will simply disappear rather than offer work.

The rotors have shallow channels of sorts in them,

I'd like to see one run on E85 though, and with a bit of boost. Should make for interesting numbers.

They are reasonable sized combustion chambers and most of them end up about 9:1 from memory. Racing Beat tried to sell 10:1 high compression, lightweight rotors just the same as you would buy high comp, lightweight pistons but oddly sales were very poor and they stopped making them.

### #5 bigleagueslider

bigleagueslider
• Member

• 1,117 posts
• Joined: March 11

Posted 24 March 2012 - 05:53

The max basic geometric CR (Vmax:Vmin) for a classic Wankel with a tri-lobe rotor, 2 lobe epitrochoid housing, and a 3:2 ratio phasing gear, is only about 12:1. But that value can vary based on the actual dimensions used to generate the epitrochoid geometry. Most SI Wankels use side intake and peripheral exhaust ports to minimize overlap.

### #6 gruntguru

gruntguru
• Member

• 5,992 posts
• Joined: January 09

Posted 24 March 2012 - 06:35

In the blue corner at 20:1 - Cheapracer
In the red corner at 12:1 - bigleagueslider
Gentlemen touch gloves . . .

### #7 Catalina Park

Catalina Park
• Member

• 6,251 posts
• Joined: July 01

Posted 24 March 2012 - 06:48

### #8 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 24 March 2012 - 07:06

In the blue corner at 20:1 - Cheapracer
In the red corner at 12:1 - bigleagueslider
Gentlemen touch gloves . . .

Anyone with practical hands on experience with engines only has to see the sizable combustion chamber area in a rotor (Mazda at least) to know that 12:1 being the limit is nonsense, however it's not relevant anyway as increases do not offer any gains, that has been proven both because of heat path and something hitting the back of my brain about flame path needing a reasonable amount of clearance or you get incomplete burn - in other words the combustion chamber in the rotors is as much a flame path as it is volume control.

That would also explain why Mazda Rotary's don't respond to compression increases but contrarily respond to pressure charging well.

I ain't no gentleman either, a swift kick in the balls will finish it.

### #9 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 24 March 2012 - 07:07

### #10 Catalina Park

Catalina Park
• Member

• 6,251 posts
• Joined: July 01

Posted 24 March 2012 - 08:24

I always understood the size and shape of the combustion chamber to be about flame path and complete burn and not about compression ratio.

### #11 GeoffR

GeoffR
• Member

• 516 posts
• Joined: January 04

Posted 24 March 2012 - 11:58

I will admit to being a technical ignoramus in regard to a lot of subjects discussed on this forum, but this one interests me as being a Mazda rotary 'user' many years ago. In regard to this subject, how would you increase the compression ratio of a rotary engine, by way of a reprofiled rotor? Would I be right in assuming that that would be similar to putting high compression pistons in a conventional engine? Is there anyone who makes or modifies rotors to increase the compression ratio of rotaries? The one thing that I really loved about rotaries was their ability to keep revving, which was really handy in rallies where there were shortish straights between tight corners, just keep it in one gear!

### #12 MatsNorway

MatsNorway
• Member

• 2,239 posts
• Joined: December 09

Posted 24 March 2012 - 22:25

Static comp is one thing. But how does its filling percentage compare to a piston engine?

I guess its a rpm related issue. and since according you you guys Wankels doesn`t like high CR it explains its fairly peaky power curve tendencies. Only comparably good at high revs.

### #13 bigleagueslider

bigleagueslider
• Member

• 1,117 posts
• Joined: March 11

Posted 26 March 2012 - 00:54

Anyone with practical hands on experience with engines only has to see the sizable combustion chamber area in a rotor (Mazda at least) to know that 12:1 being the limit is nonsense, however it's not relevant anyway as increases do not offer any gains, that has been proven both because of heat path and something hitting the back of my brain about flame path needing a reasonable amount of clearance or you get incomplete burn - in other words the combustion chamber in the rotors is as much a flame path as it is volume control.

cheapracer,

Here's a link to the definitive text on the Wankel engine by Kenichi Yamamoto:

www.iwankel.com/manuals/.../...

The calculations for max theoretical compression ratio are on pages 27 & 28. If you know the design parameters for a Wankel, you can calculate the max theoretical geometric CR.

I will agree with you that getting a theoretical geometric (Vmax:Vmin) CR above 12:1 is possible, but it is not practical with the classic 4-stroke SI Wankel rotary. As you noted, the classic Wankel uses a dished rotor face, and this is necessary to create a suitable combustion chamber shape. Due to considerations of combustion chamber shape, SI Wankel engines don't normally use geometric CRs above about 10:1. In fact, even Mazda's R26B 4 rotor race engine that won LeMans in 1991 was only 10:1.

www.rotaryeng.net/Mazda_R26B_US.pdf

As MatsNorway notes Wankels perform well at high speeds, due to high intake charge inertias, no valvetrain components, and excellent dynamic balance.

Lastly, I'd appreciate it if you'd refrain from kicking me in the family jewels.

slider

### #14 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 26 March 2012 - 03:18

Static comp is one thing. But how does its filling percentage compare to a piston engine?

I guess its a rpm related issue. and since according you you guys Wankels doesn`t like high CR it explains its fairly peaky power curve tendencies. Only comparably good at high revs.

Tried to avoid them most of my life but I actually have experience with them because I helped prepare a very well known one in Oz Rallying (Hugh Bell's Pedders Dazda and RX7) and also my workshop was across the road from Reliance Automotive in Canberra for 2 years being good mates with John Waterhouse (6 rotarys at the last year of Group C Bathurst were John's engines for example).

Their power curves is no more peaky than a piston engine and like any engine comes down to exhaust and induction timing, lots of it means peaky but mild porting is that, mild.

I tried to convince John to try reed valves on his 260hp rally 13B periphial port engine to broaden the torque curve but he never did - John was Canberra rally champion a few times and he had a bridge port or PP engine choice depending on if the rally roads were open or tight, I believed the reed valve mod would have covered both.

### #15 gruntguru

gruntguru
• Member

• 5,992 posts
• Joined: January 09

Posted 26 March 2012 - 03:53

Reed valves sounds like a great idea. Have they been used successfully at the frequencies required for a rotary (with three chambers sharing the same port)?

### #16 Lee Nicolle

Lee Nicolle
• Member

• 7,481 posts
• Joined: July 08

Posted 26 March 2012 - 05:32

Brap brap brap,,,,, brap brap brap.
Noisy antisocial horrid things. They are a little more peacefull with a hairdryer on them.
20Bs are a little less harsh on the ears but still noisy.
In motorsport they seem to be just about dead, 808s are gaining piston engines instead of chicken cookers.
The noise has killed them really, like some other engines if you make them quiet they lose any advantage they had.
Though as a standard road car engine they are quite fun though quite thirsty.

### #17 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 26 March 2012 - 06:16

Reed valves sounds like a great idea. Have they been used successfully at the frequencies required for a rotary (with three chambers sharing the same port)?

Yes, nothing for millions of 2 strokes to rev over 10,000rpm every weekend with them for the last 40 years - they also open and close twice each rpm so you're looking at 20,000+ cycles with no issues at all. A common 250cc MX engine sold in the millions is "flowing" 50+hp through them.

A Rotary would only be one cycle per RPM at low to mid speeds and generally only open at high speeds.

I have no idea why no one has done it including Mazda themselves, seems totally logical for a Rotary especially if you have ever witnessed the fuel mist above a 48IDA Weber on a PP or Monster Port.

Take it away Christy ...

On that note I don't understand why no exhaust port valve (changes height hence timing of exhaust port) that also millions of 2 strokes use successfully every weekend.

### #18 gruntguru

gruntguru
• Member

• 5,992 posts
• Joined: January 09

Posted 26 March 2012 - 06:29

Yes, nothing for millions of 2 strokes to rev over 10,000rpm every weekend with them for the last 40 years - they also open and close twice each rpm so you're looking at 20,000+ cycles with no issues at all. A common 250cc MX engine sold in the millions is "flowing" 50+hp through them.

A Rotary would only be one cycle per RPM at low to mid speeds and generally only open at high speeds.

A two stroke also has only one induction per revolution per cylinder - not two.

The Wankel reed valve would need to be a bit larger and heavier - flowing over 100 hp per reed.

Good point on the valve being constantly open at high rpm.

### #19 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 26 March 2012 - 07:03

A two stroke also has only one induction per revolution per cylinder - not two.

Reed valved 2 strokes since about 1975 have the main induction through a piston induced crankcase low pressure that most are familiar with (2 stroke loop induction 101) and then an exhaust sonic wave induced induction late in the cycle directly through a boost port (usually just past and above the reed) entirely bypassing the inlet/crankcase/transfer ports that many are not familiar with - this causes the reed to open twice.

Actually and I'm bloody amazed, this gif actually shows the sonic wave/boost port in action and the 2 distinct reed valve actions ..

Step by step here http://www.southerns...rokeengine.html

Mysterious things are those "simple" 2 stroke engines.

Edited by cheapracer, 26 March 2012 - 07:06.

### #20 gruntguru

gruntguru
• Member

• 5,992 posts
• Joined: January 09

Posted 26 March 2012 - 07:17

Awesome! Thanks Cheapy - although I can't actually see the secondary induction path I can sort of visualise it.

### #21 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 26 March 2012 - 08:43

although I can't actually see the secondary induction path I can sort of visualise it.

Well there's one long blue fuel mist and a short blue 'puff' in the .gif above - concentrate on the path of the short puff as it happens ..

Direct route - carb to reed to boost port to combustion chamber.

### #22 manolis

manolis
• Member

• 606 posts
• Joined: May 03

Posted 26 March 2012 - 09:06

At Wankel_Compression_Ratio_Limit it is a windows - exe program (made in Visual Basic) that calculates the compression ratio limit of the Wankel Rotary.

The RX8 Wankel (eccentricity 15mm, rotor radius 105mm, chamber width 80mm), with 654cc per chamber, has a theoretical maximum compression ratio of 18.3.
In practice it drops below 17 due to the necessary gap between the parts (thermal expansion, bearing clearance, manufacturing accuracy of the parts, space for the sealing "rings", clearance in the internal-external gearing, bending of the power shaft etc).

Besides, the poor sealing (one only sealing "ring" per apex, line - and not surface - contact of the sealing "ring" with the "casing") leaves a good amount of compressed gas to escape to the other two chambers of the same rotor.

Lowering the eccentricity relative to the rotor radius, the compression ratio increases (even above 20:1), the shape of the combustion chamber worsens, the capacity of the engine - for same external dimensions - lowers.

The lower the "step" scroll-bar, the slower and more accurate the calculations.

For slower PCs the "step" scroll bar needs to be at top end.

Thanks
Manolis Pattakos

### #23 manolis

manolis
• Member

• 606 posts
• Joined: May 03

Posted 28 March 2012 - 03:37

Helloooo...

Is anybody there?

I understand the "not responding" policy, but I cannot understand being afraid to download and open a program (the statistics of pattakon web site say so).

For my last four "strictly technical" posts:
Cat-and-mouse rotary engine / flyer
2-stroke uniflow with true four stroke lubrication
Constant velocity joint
Wankel compression limit

the true responses were from none to zero, while in the topic "Enough" (regarding the banning of a member of the forum, now deleted) 91 replies were posted!

Am I banned, too, or are these topics "too technical" to deal with?

Thanks
Manolis Pattakos

### #24 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 28 March 2012 - 04:42

Hi Man.

I offered my opinions in posts #4 and #8 above on Wankel compression ratios.

I like your ideas on the 2 stroke but see no cures for the typical combustion troubles that are stumping everybody (which is handy for some as they can continue their "investment grabs...") mind you for a massive marine engine that you're alluding as being your market then that probably doesn't matter.

I have no need for a Wankel compression ratio program but thanks for offering it

I love your website, often drop in to see what's new - keep up the brain storming!

Edited by cheapracer, 28 March 2012 - 05:03.

### #25 manolis

manolis
• Member

• 606 posts
• Joined: May 03

Posted 28 March 2012 - 11:03

I put the PatMar(ine) in a new thread to talk why it is limited to large marine and power station engines.

Manolis Pattakos

### #26 MatsNorway

MatsNorway
• Member

• 2,239 posts
• Joined: December 09

Posted 28 March 2012 - 15:31

Slight OT about the two strokes. The serious race engines don`t use reed valve but a slot in the crank. and i think some have a rotating valve system.

### #27 Tony Matthews

Tony Matthews
• Member

• 17,503 posts
• Joined: September 08

Posted 28 March 2012 - 16:33

a rotating valve system.

Disc valves.

### #28 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 28 March 2012 - 17:29

Slight OT about the two strokes. The serious race engines don`t use reed valve but a slot in the crank. and i think some have a rotating valve system.

There are 2 main types for motorcycle road racing, one is a disc valve that you are probably referring too (rotating valve) that has a rotating thin disc with a cutaway to control the intake timing and reed valves.

A disc valve will have slightly more top end having unrestricted but set (asymmetrical) port timing while a reed valve enjoys more mid range and throttle response but the reed petals and block itself offers a slight flow disturbance but there's not much in it and it's very track dependent as to who's favour it's in. 250cc GP in their prime years had Aprillia with disc valve and Honda/Yamaha with reed and it generally balanced out with tight and open tracks throughout the years.

Max Biaggi hopped off an Aprilia after winning the World Championship, hopped onto a Honda the next year and won it then hopped back onto the Aprilia the next year and won it again to show how close they were.

All have some sort of exhaust port valve that controls exhaust port height which controls port timing, some are based on a rotating barrel and some are blade type - they all achieve the same result just they are all patented to their respective manufacturer so they have to use different designs.

For motocross and any off road dirt bike use, reed valves can't be beaten for their power spread and throttle response which is more important than a couple of % outright power.

In 1975 PUCH made a 250cc MX bike that won the world championship that had 2 carbs, one a normal old style piston port and the other fed a disc valve - even to this day it is one of the most powerful production MX bikes ever made with over 50hp off the floor. A tiny little Austrian scooter company took on the best of the Japanese and beat them in the world championship! Ironically one of the few companies to ever beat the Japanese again was also Austrian, being KTM of course.

http://www.earlyyear...uch2501975.html

Edited by cheapracer, 28 March 2012 - 17:33.

### #29 John Brundage

John Brundage
• Member

• 296 posts
• Joined: July 08

Posted 28 March 2012 - 19:18

Helloooo...

Is anybody there?

I understand the "not responding" policy, but I cannot understand being afraid to download and open a program (the statistics of pattakon web site say so).

For my last four "strictly technical" posts:
Cat-and-mouse rotary engine / flyer
2-stroke uniflow with true four stroke lubrication
Constant velocity joint
Wankel compression limit

the true responses were from none to zero, while in the topic "Enough" (regarding the banning of a member of the forum, now deleted) 91 replies were posted!

Am I banned, too, or are these topics "too technical" to deal with?

Thanks
Manolis Pattakos

I like your website. I need a little free time at home to explore the programs. I am eager to try them out.

### #30 MatsNorway

MatsNorway
• Member

• 2,239 posts
• Joined: December 09

Posted 28 March 2012 - 21:15

In 1975 PUCH made a 250cc MX bike that won the world championship that had 2 carbs, one a normal old style piston port and the other fed a disc valve - even to this day it is one of the most powerful production MX bikes ever made with over 50hp off the floor. A tiny little Austrian scooter company took on the best of the Japanese and beat them in the world championship! Ironically one of the few companies to ever beat the Japanese again was also Austrian, being KTM of course.

http://www.earlyyear...uch2501975.html

Why don`t they make such designs anymore?

The nitro powered RC engines only got a slot in the crank as far as i know. 40 000 rpms..

http://www.rc-trucks...tro-engines.htm

Here is a fun one.

### #31 Johan Lekas

Johan Lekas
• Member

• 52 posts
• Joined: May 09

Posted 29 March 2012 - 09:22

Helloooo...
Is anybody there?

I read all your post with interest, but often feel I don't have experience or knowledge to contribute with anything of value

You obviously have a great ability to free yourself from the conventional thinking around a subject and find truly new workable solutions (that are just not different only to aper novel)
It's great to follow!

### #32 manolis

manolis
• Member

• 606 posts
• Joined: May 03

Posted 29 March 2012 - 14:01

I read all your post with interest, but often feel I don't have experience or knowledge to contribute with anything of value

You obviously have a great ability to free yourself from the conventional thinking around a subject and find truly new workable solutions (that are just not different only to aper novel)
It's great to follow!

Hello Johan ( GIA HARA APO ELLADA ).

Anybody unfamiliar with a subject has the advantage of the “fresh-eye”. The rest are mostly "luck" and "persistence".

This was the case with the MultiAir.

A big OEM ( the FIAT group ) made a great invention that reliably turns the ICE into a digital device.

Think for a moment about the MultiAir from a different viewpoint:

It doesn’t need preparation or motion of mechanical parts.
The ECU sends an electronic pulse “the right moment” to switch on and off the solenoid valve that modifies the valve motion (late opening, early closing, multi-lift, full lift deactivation).
The “right moment” is independently defined for each cylinder, according the program and the feedback from the sensors (high speed oxygen sensor etc).
If you want to retard the “right moment” by three millionths of a second in the third cylinder only, it is OK. The ECU is adequately fast to manage it immediately.
The ECU operates at accuracy and speed impossible for the mechanical systems.
Comparing the valvetronic of BMW to the MultiAir of Fiat is like comparing the old mechanical distributors (those with the “centrifugal” control) with the modern electronic ignition systems.

With the MultiAir the internal combustion engine works like a PC or a laptop.
It took almost a decade and a bigger than FIAT manufacturer from Germany, the Schaeffler-INA, in order to become the MultiAir reliable and cheap and ready for mass production.

When we first saw the intake cam-lobes of the camshaft, we couldn’t believe that two partners of such a size teamed up to come up with such a blunder.
We tried to communicate with them, but “no response” is their policy.
We filed a couple of patents and a prototype Alfa-Romeo car was prepared as proof of concept ( http://www.pattakon....ttakonHydro.htm )

The PatAir is better not because we say so, but because it offers more.

Toyota Prius, Mazda SkyActive-G and Nissan Micra DIG-S are three of the most fuel efficient modern engines; all three of them operate on a "limited Miller cycle" and high compression ratio; with conventional technology they achieve at least comparable fuel efficiency with that of the high tech digital MultiAir system of Fiat.

The PatAir offers more than both worlds (i.e. the high compression Miller and the MultiAir) together.

Thanks
Manolis Pattakos

### #33 cheapracer

cheapracer
• Member

• 10,388 posts
• Joined: May 07

Posted 04 April 2012 - 05:26

Some excellent wankel animations and info from the wankel masters themselves here ...

http://www.mazda.com...rit/rotary/16x/

.. a great ability to free yourself from the conventional thinking around a subject and find truly new workable solutions (that are just not different only to aper novel)

That's nice but got very little to do with it, you need a product that manufacturers are going to say "Sure, we'll drop our billion dollar investment in making our engine so we can make yours, pay you lots of money and have something the public might reject because that's good business and our shareholders will be delighted we take such risks ..."

The history of the Mazda Rotary (from the Corporate view, not the technical side) should be a "compulsory reading" directive from patent offices before you are allowed to lodge a new engine patent application.