What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

New engine.

Personally the thought of a radial with EFI gives me the giggles. I'd love to pair it with either a slicked-up composite traveling airframe or a polished metal one with retro styling (but still good drag reduction).

Exactly. I have a few people that might help me out with the EFI situation, but being odd number, and the way it fires makes things interesting with COTS systems I've talked to.

I'm doing electronic ignition for sure, but because I'm cheap it's probably going to have a carburetor on it to start.

I had quite a conversation with my GM friend about how to do EFI. I could make it "stupid" pretty easy... but it include what most people would expect with o2 sensors, knock sensors and all that is definitely outside of my scope at the moment.
 
Last edited:
Electronic ignition and EFI on a radial is complicated by not only the obvious reason of an odd number of cylinders.

Because of the geometry of a master-rod--connecting rod system, the timing of top dead center is not evenly spaced between all the cylinders. Valve timing from a ring cam, and ignition timing from a gear-driven magneto or a crankshaft position sensor have significant variations in timing on each individual cylinder.

I suppose a magneto could be customized so that the lobes that open the breaker points are positioned to make the timing on each cylinder the same.

It would be fairly easy to make an optical ignition trigger with a geared rotating disc with slots that was tuned to position the slot for each cylinder in the correct position to make the ignition timing on each cylinder the same.
 
Electronic ignition and EFI on a radial is complicated by not only the obvious reason of an odd number of cylinders.

Because of the geometry of a master-rod--connecting rod system, the timing of top dead center is not evenly spaced between all the cylinders. Valve timing from a ring cam, and ignition timing from a gear-driven magneto or a crankshaft position sensor have significant variations in timing on each individual cylinder.

I suppose a magneto could be customized so that the lobes that open the breaker points are positioned to make the timing on each cylinder the same.

It would be fairly easy to make an optical ignition trigger with a geared rotating disc with slots that was tuned to position the slot for each cylinder in the correct position to make the ignition timing on each cylinder the same.

One of the key things in my design... probably the oldest part of my design... is when I first drew the "core geometry" way back when I discovered that TDC issue. There are very few documents that explain that it's an issue and only one that I found that mentions a solution. most of the "small" pre-war engines didn't really worry about it... but the bigger and rougher they got, they had to figure out how to smooth them out. I imagine it was considered a trade secret. One of the solutions is still an active patent held by Pratt and Whitney. Thankfully I found another way. Anyway I messed with it long enough until I found a "graphical" solution for the relationship of crank angle to each piston TDC... that by itself took months before I even attempted to draw a single part. Honestly I imagine it's going to be ridiculously smooth for a radial. @scsmith you scared me for a second, I had just landed for lunch when I read you post and thought... I did fix that right... I just checked it... and yes I did. That's not an issue. My plan was to have 1... maybe 2 magnets for a HES on the cam plate. In the software, I can fine tune each pistons spark advance for tuning, then have a global advance that's pilot controlled.
 
One of the key things in my design... probably the oldest part of my design... i My plan was to have 1... maybe 2 magnets for a HES on the cam plate. In the software, I can fine tune each pistons spark advance for tuning, then have a global advance that's pilot controlled.

By putting the Hall sensor magnet on the cam rather than the crank, you can keep track of which cylinder is actually firing. Not a wasted-spark system. That is clever.

To fix the cylinder timing issue, can you design the master rod so that the rods attach at slightly shifted locations? I suppose that with a 5 cylinder there is more room to do that. On a 9 cylinder, the connections are pretty crowded.
I have an R-2800 master rod, and the bushing holes for the other 8 rods look pretty equally spaced, at least by eyeball.
 
By putting the Hall sensor magnet on the cam rather than the crank, you can keep track of which cylinder is actually firing. Not a wasted-spark system. That is clever.

To fix the cylinder timing issue, can you design the master rod so that the rods attach at slightly shifted locations? I suppose that with a 5 cylinder there is more room to do that. On a 9 cylinder, the connections are pretty crowded.
I have an R-2800 master rod, and the bushing holes for the other 8 rods look pretty equally spaced, at least by eyeball.
I'm not 100% about the R2800 but I bet they are equally spaced, It's the distance from the CL bore that is probably shifted. I did both. I wasn't about to have different length rods, and pistons with different wrist pin height.

Oddly enough... I've drawn geometry for a 3, 5, and a 7. The bigger they get... it seems like the less shift there is. Comparatively the 5 case and rod is "bigger" than a 7.... edit: I suppose that makes sense since the spacing is smaller.
 
Last edited:
I'm not 100% about the R2800 but I bet they are equally spaced, It's the distance from the CL bore that is probably shifted. I did both. I wasn't about to have different length rods, and pistons with different wrist pin height.

Oddly enough... I've drawn geometry for a 3, 5, and a 7. The bigger they get... it seems like the less shift there is. Comparatively the 5 case and rod is "bigger" than a 7.... edit: I suppose that makes sense since the spacing is smaller.

I'll look again - it seemed to me they were the same angular spacing and the same radial distance from the centerline of the crank throw. I guessed they just accepted the firing unevenness and took the dynamic loads out with the tuned-mass dampers.

The primary parameter of interest is the ratio of the stroke to the radius at the "slave" rod bushings.
 
I'll look again - it seemed to me they were the same angular spacing and the same radial distance from the centerline of the crank throw. I guessed they just accepted the firing unevenness and took the dynamic loads out with the tuned-mass dampers.

The primary parameter of interest is the ratio of the stroke to the radius at the "slave" rod bushings.

It might not be... that's a pretty "early" engine.

The tuned mass "pendulum" damper was going after a 3rd order vibration they could measure but couldn't make out where it was coming from... at least that's what was published in I think 1953. An article for the final rev of the R-3350, which to my understanding was the ultimate development of commercial radials.

I think it was a superposition event based on multiple slave links kidney bean paths slider portion masses making a transient resonance that sent the crank into torsional vibration. That's a lot to figure out by hand. They couldn't tie it to a specific ignition event or cumulative gas force as they had that plotted out REALLY well. The period didn't match up with any known motion in the engine.

I run a full speed motion analysis on my system and I don't see it... but I'm also not accurately simulating gas forces. I'm just doing a constant shaft speed dynamic analysis.

Ratio of stroke to the kidney bean "radius" I've maintained a maximum equal to that of the O-320 where the pistons came from. I didn't want to side load them any more than they were designed for. I was trying... and probably could have kept the diameter the same as a lycoming width, but that angle fought me so I think I'm like 2 inches wider. Also I kept the slave pins the same as the wrist pin diameter for now... that definitely added a bit to the overall diameter.
 
Last edited:
Why are you trying to reinvent subaru EJ 2.5 engine? With SDS , STOCK ignition,fuel injection, they range in hp from 150-185 without crazy mods! Only
thing needed, good oil & water cooling ! Tom
 
Why are you trying to reinvent subaru EJ 2.5 engine? With SDS , STOCK ignition,fuel injection, they range in hp from 150-185 without crazy mods! Only
thing needed, good oil & water cooling ! Tom
He's got a point. Aside from the number of cylinders, air cooling instead of water, the fact that it's a radial and not a boxer, it's direct drive and won't need a PSRU, a couple dozen HP, it's carbureted, and a few other minor details, they're almost exactly the same.
 
There's one in particular from 1942 discarded from the Christler design library that's basically a total design analysis of a Wright Cyclone they were under contract to build.

is there an ISBN & title/author for this publication, or was it an internal document only?

One of the solutions is still an active patent held by Pratt and Whitney.

Interesting. Suggests P & W was still working on this as late as 1998 - 99 or later?

BTW, re: the Warner heads, there are a number of good aluminum foundries in PA.

Thanks!
smt
 
is there an ISBN & title/author for this publication, or was it an internal document only?

Interesting. Suggests P & W was still working on this as late as 1998 - 99 or later?

BTW, re: the Warner heads, there are a number of good aluminum foundries in PA.

Thanks!
smt

Internal only.

I know there are a few companies making new parts... one is even making new cases. I'm curious if it's a license issue.

The alloy throws a red flag... it's only used on aircraft air-cooled engines, motorcycle, and Briggs and Stratton engines... since it's obviously not a lawnmower or Harley when ya RFQ... invariably a high up in the company calls and asks what aircraft engine you working on... and basically turn down the work.

To be fair it's a difficult alloy to cast from my understanding for people who usually just do like A356 as this stuff is nearly 200*? hotter and viscous.

I'm avoiding the entire cylinder head issue all together at the moment... just buying COTS. If I really got into it Mercury Marine actually has some of the most advanced aircooled engine alloys on the planet and I have a few contacts over there.
 
Internal only.

...when ya RFQ... invariably a high up in the company calls and asks what aircraft engine you working on... and basically turn down the work...

The correct answer to such a call is always "Custom off-road recreational vehicle engine."

Skylor
 
WrHbVrMm.jpg


I got a care package and made some new friends today!

I have 2 serviceable cylinders and a CAM that's getting sent to laser scan tomorrow to verify timing.

One is a 320 jug and the other is a 360 jug.

Apparently 360 jugs have a lower specific loading, and last longer... can typically be rebuilt. Even though I wasn't going for a 450cu inch motor... that might be what I end up with... with not much weight penalty.

What do you guys think? 400 cubic 187hp or 450 cubic inch 225 hp? It would give ya some get off the ground umph then throttle back and cruise at a pretty low power setting keeping things calm?
 
Last edited:
The alloy throws a red flag... it's only used on aircraft air-cooled engines, motorcycle, and Briggs and Stratton engines... since it's obviously not a lawnmower or Harley when ya RFQ... invariably a high up in the company calls and asks what aircraft engine you working on... and basically turn down the work.

I've heard (take your pick) various stories like that from people who actually have not worked with foundries. :( If you got it directly, it probably means the foundry is not set up/familiar with the alloy & prefers an excuse to gently brush you off. Even in that case, they will probably refer you to someone who does cast it. Is the magic elixir A201/A206? There are more than a few foundries that advertise as specializing in it. That said, I have made patterns, had low-tech products commercially cast, and sold castings in both AL & CI. I don't personally have aircraft or milspec experience. However, IME, there are competing options from experienced sources who want to cast your product, if it is a realistic project. (good tooling, realistic $)

Internal only.

There may be more than a few of us interested in purchasing a copy if you decide to reproduce it. :)

Interesting project, hope you progress.

smt
 
Last edited:
OK, I know I'm not the only one reading this and thinking, "Wow, isn't it great to be able to see this conversation about advanced engine design challenges here in an open forum?"

After reading all these pages I realize I'm still in the baby steps stages of understanding piston engines. I'm finding this discussion fascinating and highly informative. My thanks go to all here who are contributing to my continuing education!
 
OK, I know I'm not the only one reading this and thinking, "Wow, isn't it great to be able to see this conversation about advanced engine design challenges here in an open forum?"

After reading all these pages I realize I'm still in the baby steps stages of understanding piston engines. I'm finding this discussion fascinating and highly informative. My thanks go to all here who are contributing to my continuing education!

Cool! Glad you're enjoying it!

The amount of research that's gone into this is INSANE. The ability for me to pick and choose my day job projects that bleed over is pretty cool too. I'm trying to do it right.

---

I have to say, I just updated the model so that it uses the bigger 0-360 jugs. So... 450 Cubic inches, and should comfortably make 225-240 hp @2700 rpm. At the moment it weighs 171 pounds.

I see that an O-360 A1A is listed as 258 pounds... I'm pretty enthusiastic I'll be under that.
 
I am following this with great interest. I am of the same mindset as you. I used to restore old Porsche air cooled engines and am thoroughly offended by the cost of virtually identical Lycomings. Did you know that most shops get $200-250 each, for roller tappets that cost about $3-5 to manufacture?!?!

I would love to see this engine work.

I have extensive contacts in the foundry and metal casting industry so if you ever need assistance finding someone to cast something, let me know. Steel, iron, aluminum, titanium, bronze... you name it and I probably ?know a guy?.
 
I am following this with great interest. I am of the same mindset as you. I used to restore old Porsche air cooled engines and am thoroughly offended by the cost of virtually identical Lycomings. Did you know that most shops get $200-250 each, for roller tappets that cost about $3-5 to manufacture?!?!

I would love to see this engine work.

I have extensive contacts in the foundry and metal casting industry so if you ever need assistance finding someone to cast something, let me know. Steel, iron, aluminum, titanium, bronze... you name it and I probably “know a guy”.

Exactly.

yunDSBHm.jpg


I know exactly how much they cost!

The one on the left is a Lycoming, the one on the right is a GM.
They have the same bearing in them. The roller looks IDENTICAL to the point I bet they were made with the same program. The OD is identical Ø.8423... I mean seriously that's a pretty big coincidence.

I'm assuming someone over at Lycoming started with a GM and had to stretch it a a bit to get it to work with existing cases.

That said... with the lifters... My "nose" is needing some internal passageways for the hydraulic lifters. It's looking to be one of the only castings I need unless I can think my way out of it. I'm very interested in shops that are willing to do low rate projects. There is 1 place here in town that has done pretty good work for me using 3d printed wax investment casting, but they are limited in size.

You have any good design pointers for how to go about a traditional sand casting with internal passageways?

----------

I spent a solid half hour talking with a fella over at Haltech. I think I'm headed down the path with EFI port injection and electronic ignition.

I had almost resigned myself to buying a few slick magnetos to see how to cannibalize them to redo the distributor... but I don't think that's the right path forward anymore. Maybe it will come down to that eventually but I'm going to try the new modern electrical pixies because honestly that makes things REALLY simple mechanically which to quote Elon Musk here recently... the best part is no part. I'll aim for " or have an ignition system of equivalent in-flight reliability."

Besides at the price point, their entire system costs less than 2 mags.

We were talking about Oil temp, oil pressure, EGT for each cylinder, MAP and a throttle position sensor for the intake, and a crank position sensor.

So what's the cheapest autopilot with auto throttle!? LOL!
 
Last edited:
Funny I saw this today. I wish you luck!!

I saw a video on youtube last night that was a guy hand building a radial, 7 cylinder airplane engine. It was made using VW cylinders machined to suit the project. Many other parts were machined in-shop. I say it was an airplane engine because when it was started for the 1st time in the video it had a small wood prop on it. The video was from around 2013 for some reason. There weren't many videos on the channel and no more current update videos of the engine, but it was very impressive.
 
There are a few Radials on the market.

The East Europe M14 is a bit of a monster (too much for RV's) at 360HP. You can buy them new but expensive. Engine from Czechoslovakia
https://en.wikipedia.org/wiki/Vedeneyev_M14P

The Rotec (not Rotax) Radial (about 150 HP) from Australia, has developed a reputation (not good see YouTube Videos). There is an RV-8 from Spruce Creek FL flying around. Did a great job with custom round cowl, but admits to lots of extra work/time and a slower plane (in part due to lower HP and no doubt cooling drag). Engine from Australia
https://www.youtube.com/watch?v=LlPtMVijIzs
https://www.youtube.com/watch?v=ZIjh4o9QjtY

Verner makes small radials up to 110 HP. Seen some on ultralight or LSA. Engine from Romania
https://www.experimentalaircraft.info/homebuilt-aircraft/verner-aircraft-engines.php

http://vernermotor.com/

Jacob R-755 is 225 HP to 300HP and 500 lbs...is one of the smaller (not light) radials I can think of besides the Warners
https://en.wikipedia.org/wiki/Jacobs_R-755

"Warner (Scarab) 185 radial engine for the experimental airplane market was announced to be in production again.. [Original] Warner 185 rated at 200 hp for takeoff and 185 hp constant at just 2175 rpm, weight 344lbs. Engines will use today’s modern technology, better materials, higher compression and fuel octane, fuel injection and electronic ignition... Production is expected to begin in the fall of 2017, after AirVenture." (Ref Source Kit Plane 2016, status?)
http://www.warnerenginecompany.com/AboutWarnerEngines.html
https://en.wikipedia.org/wiki/Warner_Scarab


Making 200HP with such a light weight would be a feat.... Radials are cool for vintage restoration and replicas, RV's not so much. C-190/195 with a Jacob would be cool.

The VW based radial is great: https://www.youtube.com/watch?v=SyRJeZ6s8uM. This is a one off engine for display I guess. You are better off with a VW engine Horz opposed 4 cyl as has been done for decades, or if Corvair Engines are you thing, they have current popularity in the smaller Zenair and Sonex planes (www.flycorvair.com)
 
Last edited:
Thanks for posting that Verner link. Beautiful plane, install and engine. Just want something of my own... maybe a few others will too.
Cool custom replica. Get building. Verner 9S is new... cool radal, 158HP and $27,000, and far better than the rotec in my opinion... The Samson replica is based on a Pitts S1. It's a serious project but looks doable. The link below is a great overview of all Verner engines including the 158HP 9 cylinder at $27,000 new, and how to buy (2 USA dealers). Bargain.

https://youtu.be/VNq5QAAR5_I
 
Last edited:
That said... with the lifters... My "nose" is needing some internal passageways for the hydraulic lifters. It's looking to be one of the only castings I need unless I can think my way out of it. I'm very interested in shops that are willing to do low rate projects. There is 1 place here in town that has done pretty good work for me using 3d printed wax investment casting, but they are limited in size.

You have any good design pointers for how to go about a traditional sand casting with internal passageways?

If you have any sort of drawing or schematic, I would be happy to take it to a couple of my foundry guys and see what they think. Sandcasting is going to be the least expensive option, utilizing a job shop, I would think. Do you have a specific alloy that you?re wanting to work with? The metal utilized narrows the options on which foundry can be used.
 
If you have any sort of drawing or schematic, I would be happy to take it to a couple of my foundry guys and see what they think. Sandcasting is going to be the least expensive option,[/quote[

Hmmmm....

Traditional wood/composite pattern on a match plate? Loose split patterns? Are you thinking green/petrobond sand, or no-bake?

Depending on number and contours, the core boxes will be the "interesting" components. Do you know any foundries that still paste up cores, or will shell core boxes be necessary?

If the foundry knows ahead of time that there will only ever be -one- cast, the rigging development and charges might be costed to his invoice.

One only, I'd figure a design that allowed machining it. Even if that required some assembly as opposed to en-bloc construction of the component.

Not naysaying, i really like pattern/foundry work. But it does not sound like the inexpensive way for one-only. Keep in mind, the casting still needs heat treating and machining, possibly twice. If the OP already had extensive pattern experience and has worked with foundries before, then go for it.
You can call your labor for the tooling (pattern & core boxes) "free" and maybe the foundry will apply the rigging at no or low cost if they are impressed with your pattern and matchplate work & find the project interesting. Good luck with it.

smt
 
I'm aiming for half that.

Good luck. Keep us informed on your progress. However a scratched built 200HP radial for $14,000 is an aggressive goal. The Verner 9s uses magnesium and titanium to keep weight down and is a bargan at the price of a Lycoming.

If your goal is to fly of course purchase an engine (Lycoming). If your goal is develop an engine then go for it. However you will not save money. Many have tried and spent decades doing it.
 
Last edited:
Right now I don't have the nose fully defined. I'm still waiting on my 3D scans to make sure I position things in the right place... mainly the angle of the pushrods, so the pushrod covers are not anything exotic into the head. Right now the position of the entire lifter assembly, thus cam gear train fore-aft is not defined yet.

I've actually done a lot of 3D printed wax/investment castings.
Prices are pretty reasonable all things considered. I've even run low production of 50 parts using 3D printed cores. I would imagine this "nose case" would be in the $1500 range based on it's size finished.

As far as alloy, I'm assuming I can get away with typical A356.0 It's temper temp is well above max oil temp. Thoughts?

I'm really not familiar with traditional sand cast. Most everything I've done professionally... thus my contacts are investment castings.

One thought on keeping it machined would be to have the oil "ring" that supplies the lifters be on the mating face. It would put about 780 pounds of force on the mounting hardware which isnt much... after all that hardware is literally pulling the airplane through the sky, but I'm concerned about asking a gasket to do that. It's well within the scope of a product I like called threebond 1215... but I don't know how I feel about doing that, reminds me of a water pump gasket on a car... and that still leaves the main nose bearing dry. Looking at a lot of turbines... PT6 is coming to mind, I'm not opposed to external oil lines, but then that's fittings and lines and makes things not clean on the outside.

I really enjoyed the verner engine video... after watching that I'm very tempted make the "nose" casting do as many things as I can. oil filter, oil pump... so on.
 
Last edited:
My goodness...

Finally got my part scans back! That should keep me busy over thanksgiving weekend!
EDIT: Indeed it has... I bet I know about as much about this cam as the guys who make them at this point. I needed to do this to make sure I run the cylinders exactly the same as they were originally intended.
5m11WER.jpg


So the question of the day...

Thermostatic Oil Cooler Bypass Valve... YES or Just Pressure ball and spring valves?

... Go.
 
Last edited:
Back
Top