Wankel Rotary Engine Geometry

[JCReynolds79]

Hello all,

I am doing some research on the Wankel rotary engine andd have so far determined that the shape of the housing is peritrochoidal and that the shape of the rotor is the inner envelope of said peritochoid.

I have derived formula for the peritrochoidal housing shape based on rotor radius and eccentricity, but I cannot find any info on how to create the correct rotor flank shape.

Any one able to shed any light on this please?

Many thanks,

Jon Reynolds

Regards,

Jon Reynolds

 

[EngJW]

Try to find the book by Yamamoto, it has all the equations. It might be out of print but I have seen a pdf of it somewhere on the internet. Unfortunately, I don't remember where.

There are lots of SAE papers from the Curtiss Wright days. Your research won't be complete without them.

 

[JCReynolds79]

Hi EngJW,

Ah I have that book by Yamamoto...the equations seem overly complicated. I derived my own formula for the housing in a much simpler way. Well simpler for me to understand... and I would rather be able to derive them myself than just copy them from a book.

I will continue to look around.

Many thanks,

Jon Reynolds

Regards,

Jon Reynolds

 

[TDIMeister]

Try the book by Norbye.

 

[JCReynolds79]

TDIMeister,

Are you referring to this:

https://netfiles.uiuc.edu/ro/

http://www/SocietyforExperimentalMechanics/wankelfiles/wankelgeometry.pdf

If so, I have already come across that and not found it very helpful.

There seems a bit of confusion (for me at least) as to the shape, whether it is epitrochoidal or peritrochoidal. I have been saying it is peritrochoidal as this is what is in the book by Yamamoto. But I think perhaps both solutions lead to the same shape.

As for the rotor shape, I think I can work it out. Basically by rotating the shapes around and plotting the points on the housing surface that would be closest to the rotor... hard to explain but I think I see it in my head.



Regards,

Jon Reynolds

 

[pmrobert]

Jon, have you looked at 7.5 in

https://netfiles.uiuc.edu/ro/

http://www/SocietyforExperimentalMechanics/wankelfiles/CHAPTER7.PDF

? That deals specifically with rotor flank rounding, shape, calcs, etc.

 

[EngJW]

The Yamamoto equations for the rotor flank are indeed complex and are poorly explained. Both Yamamoto and Ansdale appear to have taken them from a German book that probably can't be found anywhere. The equations do work if you program them into an excel spreadsheet. You have to use either all plus or all minus signs, and figure out by trial and error what range of angles to calculate through. They give the coordinates of a 60 degree segment.

 

[JCReynolds79]

Hmm it is still evading me... the calcs are there I know, but I want to work it out myself.

Mind, even if I wanted to use Yamamoto's equations, I am not sure how to put 'v' into the excel formulas...

v = 1/6 pi ~ 1/2 pi, 5/6 pi ~ 7/6 pi, 3/2 pi ~ 11/6 pi



Regards,

Jon Reynolds

 

[EngJW]

The v is just a number that you increment, say, from 30 degrees to 90 degrees (which is pi/6 to 3pi/6). Calculate X and Y for each value of v and you will get half of a flank. You can plot X and Y then mirror image it and rotate it 3 times, depending on what cad program you might have, and you will get the whole rotor contour.

There might be a physical significance to v but I have not figured it out.

 

[JCReynolds79]

Hi all,

Just can't get them calcs into excel properly... could well be just syntax errors as its a long line to get in one cell, with umpteen parentheses etc.

so I am putting 30 - 90 as v, but what about in the equation where it says +/- e(... should I use plus or minus??

Thanks again.

Jon

Regards,

Jon Reynolds

 

[EngJW]

Jon,

The guy at Wankel told me to use all plus or all minus. It seems to work that way.

 

[JCReynolds79]

Well I just can't get these equations to work. Nothing I have found so far is describing the method for this 'inner envelope'.

Closest I have got is a rounded triangle (right shape) but too big and rounded... this is basically a hypotrochoid. But something's not quite right.

Regards,

Jon Reynolds

 

[GregLocock]

A program that generates the Wankel rotor "envelope" has been prepared and is at

http://www.pattakon.com/tempman/WANKELN.exe

and the code is at

http://www.pattakon.com/tempman/WANKELN.txt

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[JCReynolds79]

GregLocock,

Did you do some tidying up or am I imagining replying to a post last night that has now vanished? You're making me feel like I'm going mad...

But anyway...that is the best looking description/simulation showing how it is constructed. I knew how it was made from the motion of the housing, but I just can't get my head around the maths, without looking at other examples... hopefully this one will put it to bed.

Thanks.

Regards,

Jon Reynolds

 

[GregLocock]

Manolis asked me to post the link. He's a cluey guy, contact him at his website.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.