Has anyone seen plans for a scale wankle rotary engine?

[trilect]

I'd love to build one but there isn't much in the way of information on the net. I thought about buying a plastic demonstration model and scaling but I don't that that would be accurate enough.

Any thoughts?

Similar Threads:

• 1/4 scale Case Traction Engine
• FS: scale flathead 8 model engine project
• Russian Rotary Engine (Plans)
• 1/6 Scale Offy engine project
• 1/5 scale Model A engine

[RotarySMP]

Scale of what? The Mazda 13B? How about a scale Le Mans winning 1994 26B four rotor. That would be cool.

The only plans I know of are for the SW92 from www.vth.de. I have them, the engine is 9.2cc, and was designed by Julian Faleki of Poland in the early sixties. It is all on a single A1 sized page.

Interesting plans, I haven't built it but it appears to be about as far as you can simplify the design. The Trochoid is simplified to an oval, with increased apex seal stoke compared to normal Wankels. Obviously hobby CNC was still a fair way in the future then, and the designer probably correctly figured that few would build it of you first had to build a trochoid grinder.

It is entirely of steel, with the wear surfaces to be chromed. It has no side seals or e-shaft seals, so compression is controlled by rotor clearance, as with the OS-Graubner. I bet it is a really messy engine in service, as the blow-by runs straight out through the E shaft bearings. These are standard metric bearings, so I supposed you could use seal bearings.

The rotor bearing is a roller bearing where you make the rollers which run between rotor and E-shaft.

[itsme]

HOORAY, more rotary engine interest!!!

The only plans that I've ever come across are the ones that Mark has already mentioned. I've looked for others, but have never found any.

I'm definitely going to have a go at a wankel in the near future. I just need to get the CNC mill going and then housings and rotors can be made by the dozen (well, sort of...). I'm tempted to give the VTH plans a go, but then if I've got the CNC capabilities, I don't see why I shouldn't use a 'proper' housing. It really isn't much more work, and it can all be done extremely accurately using maths. I'm still working on my spreadsheet to generate code for the housing and it seems to be working quite nicely so far - the big test will be when the machine is up and running. Maybe I'll just 'steal' some ideas from the VTH plans and design my own engine.

A 4 rotor Mazda engine sounds like quite an idea. That would certainly be a lot of fun. My ideal scale engine would have to be the original DKM 54. I love the concept of everything rotating about its own axis. In 1957 they had the first prototypes running at 25000 RPM. In fact, if I'm designing my own engine, I don't see any reason why I can't build a DKM type engine. This could be a lot of fun!!!

As for the plastic models, I wouldn't recommend that. I've got one and I'm not sure that you could copy it and end up with a working engine. They work in plastic and that's the material they should stay in... (You should still buy one though - they look good on your desk )

I just need to get on with this CNC conversion now...

Regards
Warren

PS. Wasn't the 4 rotor Mazda R26B engine from 1991???

[trilect]

I found that site but I couldn't find the plans link, Unless there is a english button my browser isn't picking up on.

[itsme]

Hi there,

Sadly I couldn't find an English version of the site either, so a bit of guess work comes in handy. Here is the link to the page where you can find the wankel plans:

http://www.vth.de/shop/shopfiles/sho...d=10102&10102=

It is called the Wankel SW92 and is towards the bottom of the page.

You can always find an online German dictionary...

Regards
Warren

[trilect]

Hi there,

Sadly I couldn't find an English version of the site either, so a bit of guess work comes in handy. Here is the link to the page where you can find the wankel plans:

http://www.vth.de/shop/shopfiles/sho...d=10102&10102=

It is called the Wankel SW92 and is towards the bottom of the page.

You can always find an online German dictionary...


Thankyou !!

Regards
Warren

Thankyou very much, For some reason bablefish interpreter isn't working for me.

Thanks again
Sam Gordon
Louisville, KY

[itsme]

Hello again,

If you are interested in buying these plans, there is a place in the USA that sells them. I think they are identical to the ones from VTH (and are still in German), but at least you can order in English (and they are already in the US)...

The site is:

http://www.sssmodels.com/

You'll need to go to the 'books, drawings, videos' link and then to the 'engineering drawings' link and scroll down a bit. They're $9.75 from there.

I hope this helps

Warren

[RotarySMP]

That english site is cheaper than VTH.de.

Sorry, it was 1991 that Mazda won Le Mans. I should have checked the date.
http://www.monito.com/wankel/lemans.html

The Wankel is a simple engine with the devil in the details. It is no suprise that it took so long to sort out all the material compatibility issues etc.

I have started designing wankels a few times, but set my design requirements too high, and you start getting a machine which look like a scaled 13B, but would be a hell of a project. The SW 92 is pretty cool as it is probably about a simple as you can build one.

[fyffe555]

As Mark says these things are very simple - but very difficult to produce. I played with norton twin rotors years ago and got facinated enough to go into them a little further.

What we call a wankel engine actually isn't - or rather it's not the first and only configuration he produced. Wankel first produced a DKM engine where the rotor was concentric to the shaft and it and the peripheral housing with the funny shape both rotated. No eccentric motion. The Engine we call a wankel rotary is a KKM eccentric rotary first produced by wankel and NSU. The material problems took 30 or 40 years to sort out.

Wankel has over 2 thousand patents on rotarys and over 2 hundred are on different shape rotors and housings. The rest are on part design, pumps, steam engines, superchargers and the like. Clever guy.

To design the basics of a kkm rotary isn't difficult - it's difficult if you want it to work! and especially if you have high design requirements and even more so when you get to the details that will actually make them work - as Mark has said.

The Housing shape is actually easily defined and drawn and so presumably turned into Gcode. It's an epitrochoid, or in english it's the shape produced by a point on a circle rolling around another circle twice the diameter of the first, much the same way as a spirograph works.

Rotor shape isn't actually very critical apart from a couple of things. The rotor cycloidal gearing turns the rotor at one third the speed of the shaft. The rotor face(s) must produce the approximate compresion ratio and is a function of the size of the housing. Approximately it's an arc of a circle center on one point of the rotor passing through the other two points.

Where it gets tricky is placing the spark plug and ports and even worse - tip seal design and rotor seals... and then you've got to get that housing surface chromed or otherwise hard plated smooth and polished...

When running nortons we had a twin rotor 588 cc (rated) at over 170hp that you could hold with one hand.

Andrew

[skippy]

I'll never forget the sound of the Mazda quad rotor WSPC (world sports prototype car?) going around Spa Francorchamps in Belgium back in 1989 (or was it 1990?) Coming out of the corners it sounded like a Cosworth V8 at full revs and then it would rev from there upwards. I know we're talking about models here but have a look at these two. Although it's from the world of dragracing which not everyone likes, they are however impressive projects.
(A) Quad rotor, huge single turbo, making around 1600 hp. Best ET (dragracing) 7.29 seconds at 191.9 mph
http://www.queenstsmash.com.au/v3/pages/racing_qsr1.php

(B) or this single 20B running 7.04 seconds at 199 mph
http://www.pacperformance.com.au/con.../Team/mx6.aspx
Phil

[itsme]

Hi there,

Producing Gcode for an epitrochoid is no problem at all. A simple Excel spreadsheet can be used along with the parametric equations for the shape.

The porting on these engines is certainly not the simplest thing to get your head around, especially when timing is quoted in degrees like a 'normal' reciprocating engine. The sealing is also a problem and that is one of the things that took Felix Wankel the longest to try and figure out. His original 'sealing grid' was far more complex than the seals that you now see on a Mazda (or modern) Wankel engine.

In my opinion, a DKM engine would not be too much more difficult to build than a KKM engine. The housing would be similar (although cooling would be difficult), the rotor would be a bit more complex (since induction is through the rotor and ignition plugs are also in the rotor), however a lot of the other parts are 'simpler' to build. There is no eccentric shaft to machine and no complex bearings within the rotor. One of the big attractions for me, is that there are no internal gears used. All phasing gears are spur gears that are essentially sitting outside of all the action. The housing would need to be balanced properly if high speeds are expected (which they should be).

Overall, the DKM is a different, more complex engine than a KKM and it isn't surprising that it was simplified for production. For specialised use though, it makes so much sense. The performance is excellent and in 'special' applications (like my model aeroplanes or workbench... ) disassembling the engine to change the plugs could be tolerated.

I definitely think that more of this little engines should be made and now with CNC machines available to model engineers, it is actually within reach. I need to hurry up and get that mill going now!!!

I wonder what ever happened to all those wankel engines that were produced in the 60s and 70s. I would love to get my hands on some of those...

Regards
Warren

[Eson]

Isme: Id love to know how you do this excel thingy in CAD, i use Solidworks but i cant get it working. I got the formulas, so thats not the problem. I got problems to get the epitrochoid house exact.

fyffe555: "The Housing shape is actually easily defined and drawn and so presumably turned into Gcode. It's an epitrochoid, or in english it's the shape produced by a point on a circle rolling around another circle twice the diameter of the first, much the same way as a spirograph works." - This isnt entirely correct because the wankel got the formula Y=e*COS(3a) + R*COS(a) , X=e*SIN(3a) + R*SIN(a). Where e is eccentricity, the amount of offset between the eccentric shaft; R which is generating radius, the distance between the rotor centerline and the rotor's apex. a is 0-360 degrees.

And those who dont know anything about Wankels, try this german page: http://www.der-wankelmotor.de/Techni...km_vs_kkm.html and then use Babelfish: http://babelfish.altavista.com/

[itsme]

Hi Eson,

The spreadsheet is very simple. All it basically does is plot lots of different points (I put 720 points to start with) using the equations and then sticks these points into G-code form. It doesn't involve any CAD programs at all. This will generate a single 'pass' or cut. In order to produce full G-code for the shape (ie. including depth or z-axis), you can just copy the code from Excel and then paste it many times in Notepad, adding in the z-axis inbetween. I'm pretty sure this could all be done in one step on Excel, but I haven't tried yet.

I don't have a CNC mill running yet, so I obviously haven't actually cut any parts using this method. I have however stuck the code into Cutviewer and it definitely looked like an epitrochoid.

If I have some time later on today, I will do a bit more work on the spreadsheet and then maybe post it somewhere in the near future.

Regards
Warren

[fyffe555]

Eson, yup, your equations are just a simplification of those to define an epitrochoid, of a ponit on one circle rolling around the other, which is what I described.. http://mathworld.wolfram.com/Epitrochoid.html

Another tip, should you try to use the spreadsheet method, and if you didn't already know, is that excel uses radians not degrees solving trig functions like cos, sin and tan. You need to convert the angles to radians or your curves will look very strange.

Once you've got the x & Y output you can plot it in excel, import the datas as a spline or poly lin ein acad or whatever, or convert it to gcode directly as suggested but I've not done that.

Andrew

[itsme]

Hi there,

The equations are just for a circle rotating around another. The ones that Eson gave have been adapted to suite the configuration of a wankel engine.

You will need to convert all angles to radians and another nice feature in Excel is graph plotting. I normally plot a XY point graph (or something to that effect) and then you can actually see the shape that is being created. When you end up with some strange curly thing, you know that you've gone wrong.

So, who here is planning on making a model wankel engine then????

Regards
Warren

[itsme]

Sorry Andrew, I'm repeating what you've already said. I missed the part about plotting in Excel in your last post...

[fyffe555]

Hey warren, no problem - I do it all the time... i'd be interested in how you get gcode out of the spreadsheet though. I've just done it and there's been lots of editing, replacing, inserting etc going on..

You guys have got me started so I think I'm going to see if I can't make up a model - not running, just platic maybe.

Andrew

[itsme]

Hi there,

I normally make a column that has all the calculated x values and a column that has all the y values. In a new column, you can then do something like this:

="G01 X "&ROUND(G2,4)&" Y "&ROUND(H2,4)&" F300"

The first part will write 'G01 X', the next part will take the value in G2 (which is an x value) and round it to 4 decimal places (you can put as many decimal places as you like) and then insert it next to the first part, the next part puts in 'Y', the next part puts in the value from H2 which is rounded again and the final part puts in 'F300' which is the feed rate. This is then copied down next to all the coordinates and will give Gcode. The gcode column can then be copied and pasted into Notepad etc. Obviously the whole thing can be customised and altered to do whatever else is needed.

As I said, I've put the resulting Gcode into Cutviewer and that seemed to understand what I was trying to do, so it does work. Andrew, let us know if you are going to make a model and post lots of pictures. It'll be good to see.

Regards
Warren

[fyffe555]

Warren, thanks that Excel routine works great and the output runs through cnc simulator fine. Even more convoluted the gcode goes through the acad import script to produce an acad polyline.

Need to polish up the formula, I've five variations on the epitrochoid theme, none of which seem to be totally accurate.

Not commiting myself to a model but I think I'll try something along the lines of an open working model. Maybe even out of mdf - like balsamans clocks which where dead impressive - and colour paint the parts.., just maybe - I've way too many projects on the go as it is..

Andrew

[RotarySMP]

The practical wankel trochoid is offset a distance outside the trochoid generated by rotating the circles. I don't know the actual range of offset distance: major diameter ratios which work.

http://www.wankel.org/epitrochoid.jpg.