Rotary Engine Con Rod Crank Pin Assy

[sbwhart]

I've been chewing arround the best way to configure the con rods on the Three cylinder Rotary I'm building.

I'm aware of three there may be more, please shout if you know of any others



1:- Is the method used by Elmer for his three cylinder rotary, what you've got is a spool that centres on the crank pin, with the con rod equaly spaced around the spool on their own crank pin.

2:- Is the method used in the Haylo, what you've got her is a master con rod attached to the crank pin, with the other two con rods attached to the master con rod.

3:- Is the method used in the Cygnet Royal what you have her is a master con rod with a extened bearing bush attached to the rod that roates around the crank pin, the other two rods run on the outside of the bearing bush with cranked con rods.


Like most things all these designs are comprimises 1:- Is compact but not very robust, and you don't have a stright action on the pistons 2:- again is compact but you don't have a straight action on two of the pistons3:- Is not as so compact but you have a straighter action.


What I've had in mind is to modiified version of option 3:-

4:- Instead of having the extended bearing bush, just have a short bush on the crank pin master con rod with the same on the cranked con rods all bearing round the crank pin:- like this




Any ideas, comments, criticism will be welcome.

Cheers

Stew

[tinkerer]

Check this out. I don't know how you coould improve on this design concept.
http://upload.wikimedia.org/wikipedia/commons/a/ab/Radial_engine.gif

[bogstandard]

Stew,

As you know, I have built the Cygnet Royale, and the crank shown on the plans works very well.

You have to be very careful when going over to another crank system, especially with shorter rods. The whole geometry of the crank swept area can change.

The only real way to do it would be to put what you propose into a CAD package and when animated, see if any obvious problems show up.

The usual problems are that the conrods can hit the bottom edges of the cylinders because they are sweeping over a larger angle, and length of strokes can change, causing either too little or too much movement of the piston, which in turn could require a cylinder length change and maybe timing adjustments.


John

[kvom]

While a CAD emulation is nice to have, you can calculate most of the critical factors yourself, and using a spreadsheet lets you experiment with various dimensions.

The stroke will be just twice the distance from the rod end to the crank pin regardless of the configuration.  The angle the rod makes is at a maximum when the crank pin is at 90 degrees to the cylinder centerline and can be calculated with a bit of trig.

A nice thing about a radial like Halo is that the input and output ports are all at the top of the cylinder, so that as long as the piston stays within the cylinder boundaries it will run.

[Russel]

Keep in mind that configuration number one, Elmers, the hub is soldered or brazed to one of the rods, so it also has a master rod. The halo version, number two, the hub and master rod are machined in one piece. 

I realize that you are probably considering the differences in machine methods required, more that the configuration of the designs. I just thought I would point out that the Elmers version has a master rod also because it isn't super apparent in your diagram. (I know it took me a little while to grasp how a radial engine could connect all the rods to one hub without destroying itself while it was running.)

Russ

[sbwhart]

Thanks for the input Guys

Russ:- I completely missed the fact the Elmer hub was fixed to the spool making it a master rod, thanks for pointing this out.

I can remember reading an article in model engineer about radial engines regarding, connecting the rods to the master and the effect on the engine timing, the article ran over a couple of issues, apparently with a full size engines the slave rods wern't spaced equally around the master rod to compensate for the geometry effects on the timing, it concluded that for scale models however the small differences involved were not significant to good running of the engine, it was these series of articles that got me interested in radials.

Thanks again

Stew

[John Hill]

I think the master rod (aka 'king rod') is the technique used in real life aero engines.

[bogstandard]

This is what I was on about, have a look at the conrod sweep on the lower left cylinder.

I don't know whether it is a drawing error or design error.

This Gif image came from the F.A.M.E site


John

[John Hill]

This is what I was on about, have a look at the conrod sweep on the lower left cylinder.

I don't know whether it is a drawing error or design error.

This Gif image came from the F.A.M.E site


John

There has to be some variation as the pins on the king rod do not trace a true circle and this is worse for those furtherest from the king rod.

[bogstandard]

John,

What I am trying to get over, on the original plans, it calls for something similar to a knife and fork arrangement on the big end, but in this case, two conrods are offset to each side of the main rod, using full length rods with a common crank pin, whereas when you go for this arrangement, with shorter rods and pivot point distances using a king rod and plate arrangement, the angles of operation of the rods change, and has to be proved before going any further as it might require modification to other components to prevents mechanical locks.

One thing I do know, having built it, the rods on this engine are very close to the cylinder bases as they are, and as I suspect, by changing the arrangement, problems will be found.

It is OK to modify things, but only if it benefits the original design, and even then, all related components have to be made compatible with the changes made.


John

[sbwhart]

John,

What I am trying to get over, on the original plans, it calls for something similar to a knife and fork arrangement on the big end, but in this case, two conrods are offset to each side of the main rod, using full length rods with a common crank pin,

Thats what proposal at 4 is.:- all con rods will rotate around a common crank pin centre line as original design.

Cheers

Stew

[John Hill]

Stew, I think your cranked rods are a good way to go as it is clear from John B. Standard's GIF that the geometry is warped a bit by using the king rod approach.  Obviously Messrs Pratt and Whitney know how to get around that!

Your cranked rods will become a challenge when you go on to your 9 cylinder version!

[sbwhart]

Your cranked rods will become a challenge when you go on to your 9 cylinder version!

Thanks John

I was thinking of a 16 cylinder Job  Like this. http://www.enginehistory.org/bristol_hydra.htm

I've watched this engine grow at model expos over the last few years.

this engine is a labour of love

Enjoy

Stew

[NickG]

Stew,

I think yours is the best concept, everything on the one crank pin an no bush because you'll be having 3 ball races the same size. They'll just have to be small ones so there is no interference due to increased big end size.

Nick

[sbwhart]

Thanks for your input chaps.

I'm going to try option 4 as this is only a three cylinder engine the cranked con rod should be OK with any more cylinders the degree of crank would become unworkable, I've got some small needle roller bearing that just mite do the job, but I've got to measure them up first and do the drawing to see what space envelope I've got to play with.

Cheers

Stew

[dsquire]

Your cranked rods will become a challenge when you go on to your 9 cylinder version!

Thanks John

I was thinking of a 16 cylinder Job  Like this. http://www.enginehistory.org/bristol_hydra.htm

I've watched this engine grow at model expos over the last few years.

this engine is a labour of love

Enjoy

Stew

I just had a look at this site and wow. Pay particular attention to the layout of the connecting rod. 5 cylinders on 1/2 of center and 3 cylinders on the other 1/2. I have attached a link to the drawing so you can see what I mean.

http://www.enginehistory.org/ModelEngines/Perkins/Hydra/HydraDwgs/HydraRods.jpg     Connecting rods for16 cylinder radial

Cheers 

Don

[sbwhart]

Hi Don thanks for the drawing link its mindbogling.

Every time I look at the site I see something new look at this for a double Rotary Table set up.

http://www.enginehistory.org/ModelEngines/Perkins/Hydra/DiffuserVanes_1.jpg

Cheers

Stew

[tinkerer]

[/quote]

I just had a look at this site and wow. Pay particular attention to the layout of the connecting rod. 5 cylinders on 1/2 of center and 3 cylinders on the other 1/2. I have attached a link to the drawing so you can see what I mean.

http://www.enginehistory.org/ModelEngines/Perkins/Hydra/HydraDwgs/HydraRods.jpg     Connecting rods for16 cylinder radial

Cheers 

Don
[/quote]
I am not seeing this. I see 8 cylinders evenly spaced at 34.8 deg all on the same plane.

[dsquire]

I just had a look at this site and wow. Pay particular attention to the layout of the connecting rod. 5 cylinders on 1/2 of center and 3 cylinders on the other 1/2. I have attached a link to the drawing so you can see what I mean.

http://www.enginehistory.org/ModelEngines/Perkins/Hydra/HydraDwgs/HydraRods.jpg     Connecting rods for16 cylinder radial

Cheers 

Don
[/quote]
I am not seeing this. I see 8 cylinders evenly spaced at 34.8 deg all on the same plane.
[/quote]

Tinkerer

Look at the spacings of the connection points on the main connecting rod. This is for 8 cylinders so there are 2 of them for 16 cylinders. The 34.8 deg dimension is to a center line not the next cylinder. 5 of the secondary connecting rods are on the right side of this center line and 2 on the left of the center line. The 8th cylinder is the main connecting rod.

Hope this makes sense Tinkerer

Cheers 

Don

[tinkerer]

Aha, I see the 34.8 deg is to the center of the slot. It depends on which center line you look at. If you look at the line from wrist pin to crank, they are equally divided, but at perpendicular center line they are divided 5 & 3. I think the other 8 cylinders would be just another bank of 8 but I could be wrong about that too.

[75Plus]

Does anyone else find it odd that the Hydra has an even number of cylinders? I cannot remember ever seeing a radial engine that had not either an odd number of cylinders or a multiple there of. The P-47 of WWII fame had two rows of 9 cylinders while the B-36, which never dropped a bomb in anger, had 6 engines each of which had 4 rows of 7 cylinders.

All the WWI rotary engines in the Air Force Museum have an odd number cylinders also.

Joe

[tinkerer]

Does anyone else find it odd that the Hydra has an even number of cylinders? I cannot remember ever seeing a radial engine that had not either an odd number of cylinders or a multiple there of. The P-47 of WWII fame had two rows of 9 cylinders while the B-36, which never dropped a bomb in anger, had 6 engines each of which had 4 rows of 7 cylinders.

All the WWI rotary engines in the Air Force Museum have an odd number cylinders also.

Joe

I did find that odd also. I have worked on several and they were either 5 or 9 cylinder. Never considered that even numbered even existed.

[sbwhart]

Hi Chaps

Domestic duties kept me out of the shed today, that and its bloody cold, well for us soft limies it is.

Spent a bit of time in the nice warm house doing a bit of CAD on the crank shaft design, until I'd drawn it out I thought it would just be a simple off set of the con rod, but no on the origional cygnet design the master rod is quite thin 1/8 but in this design this doesn't matter because of the way the bearing is orientated, but the new design requires a thicker rod to give the bearing a bit of width so you have to give the slave rods a bit more crank. Anyway after bit of fidling and head scratching  came up with this design.



I'll have to machine the con rods up with an angled web, this goes against my origonal objective of keeping the maching simple, and I've not yet worked out how to machine them.

So what do you think mmmmmmmmmmmmmmmmmmmmmmm

Stew

[Stilldrillin]

I'll have to machine the con rods up with an angled web, this goes against my origonal objective of keeping the maching simple, and I've not yet worked out how to machine them.

So what do you think mmmmmmmmmmmmmmmmmmmmmmm

Stew

Hhhmmmmmmmmmmmmm.............    

David D

[Bernd]

Stew,

I'm thinking side forces on the whole assembly. 

Bernd

[NickG]

As Bernd says, there will be a moment about the big / little ends on the slave rods but it can't really be helped, surely that is the case with any cranked rod and was so on the original cygnet design. Machining could be difficult but I know you'll find a way Stew. This may be one for the tilting head on your mill. Maybe use a bit of square stock, tilt the head over and mill the angle then you still have the datum faces for the bearing holes. Put head back square, mill down to thickness on small and big end to give you desired offset - it should work out as long as the angle is right! Might even be better to drill the holes for the bearings before milling to thickness?

Nick

[kvom]

My general impressions is that you are overthinking this, esp. for a part that won't be visible.  On my Halo build the master rod is only 1/16" thick, but all the force on it is linear  If you want more beef you could just make the master and slave rods thicker.   

[sbwhart]

Thanks for youre input chaps some good comments their, getting me back on the road.

As some of you guys said the way I was going was too complicated. So I went back to the begining and decided to see how things would look with a 1/8" thick master con rod saring the same 1/4" dia crank pin with the slave con rods to the sizes of the cygnet design.

I carfully drew up the con rods then assembled them onto a common piston to see how they would fit this showed that they interfeared with each other by about 1mm, decided to redraw taking the interference out, by reducing the thickness of the big end.

This is what I got.



Some of the sizes are a bit odd this is because I've converted from imperial to metric, when I make i'll round down to the nearest 0.1 mm with a tolereance of about 0.2. The straight webs will be quite easy to make I've worked out a what i think will be a easy method.

For the big end I've got some nice oillite bearing that i'll fit for the little ends I'll fit some phos bronze bearings.

First I'll make the crank shaft, fly wheel and pistons so that i can cut and fit to see how it all goes together.

Cheers

Stew

[madjackghengis]

Hi Stew,  as you can probably guess, I've spent some significant time on this issue before getting started on my radial engine project, and what I came up with is that for practical purposes in a scale engine, the offset of actual angular timing is insignificant, completely, and the easiest method is master rod, slave rod as most aircraft engines.  As an aside, most all are odd numbered to avoid harmonic vibrations which are multitudinous in a radial engine, and an even number multiplies them by at least twice.  One thing I've seen and looked at, is a set up pretty much as your first model shown, the one which appears to have a ring with all the cylinders free.  I have seen this, but was not aware one rod was attached, and thus it acts like a master rod set up.  I have also seen the same setup however the "ring" was substantial, with a full bushing in it, all the rods were free, and the ring was connected to the inside of the crank case with two or three straps, so it articulates with the crank while never changing its angular relationship with any of the rods.  It was among the earliest full size aircraft engines, and did fly, so it is a working design.  I think Pratt and Whitney worked out a maximum of about six degrees of difference between the two cylinders furthest apart in ignition timing, using the master rod set up, and compensated for it in both electrical ignitions and in magnetos by slight changes in the ignition cams and their angles.  Again, something that scaled down, is too small to be an impact on design.  I'd like to pursue the "sleeve on the crank with nine slave rod" idea when I get my current radial engine done, just because I haven't seen it duplicated anywhere and have only read about it.  I have seen working hydrostatic drive units that use the same concept as they need dead even power in all cylinders when working with the high pressure of serious hydraulics, and the "leashed collar" system equalizes all the angular movements for all the rods and cylinders exactly.  I just thought I'd toss that into the mix, since it is Stew, and I am building a radial, so I can get away with such a comment.  Mad Jack

[sbwhart]

Hi Jack

Thanks for your reply and insite to radial engines, I went on and built the engine using the layout in my last post and it turned out a nice runner.

Its interesting what you say about scale effects reducing significance of features, I,ve part designed OHC V4 engine, reading up on cam profiles for it the same point is made.

Cheers

Stew