Modern Aircooled 4-Cylinder Aircraft Engine Using Contemporary Engine Technology

[gtae07]

Moving things across machines faster, and moving them out the door faster, means that you need fewer machines to do the work, and less square footage dedicated to the machines and processes.

And fewer billed labor hours (which includes pay, taxes, benefits, insurance, and HR-type overhead) per part.

 

[Markproa]

.......... and you run towards the $24k...

New engines are pricey things to get ready. Risky too. Don't expect anyone to take on all of this and not charge you close to what the other guys are charging. They have to or they will go bust in short order...

Billski

Which makes the Viking engine approach pretty smart.

 

[tspear]

I get when you are designing for a million cars, that the engineering of the production, the associated high speed equipment.... That you get huge economies of scale but at a very high price point to enter the market.
This is also a highly optimized production scenario.
Exclude the drafting aspect.
Think low volume, one unit or five hundred a year. What is the fundamental difference in production costs?

Tim

 

[BBerson]

100,000 aviation engines a year will not happen.
What we need is all new plans built engines. Or partial kits.
Homebuilding is the only solution because labor is deemed free. No liability. No overhead.

 

[wsimpso1]

$12,000 means volume. GM sells a LS-3 for close to $7,000. Including engines in cars they probably make 200,000 engines. They also offset cost with parts crossover to all the other modern LS engines which is probably 6,000,000 other V8s a year. There is not 200,000 buyers, not at least short term. 50 years worth maybe. You would be good to be successful at 500 engines a year.

Other than crating up the finished engines for shipping, the LS3's are pretty much what is shipped to the factories. At Ford, we made V-8 engines for about $3000 factory cost, in the racks going to the vehicle assembly plants. No ECU, some of the wiring harness. Get into the fraction of the development, tooling, engineering, and overhead that went into each production engine, then the specialty ECU, wiring harnesses, and other near custom stuff, and I would bet there is some profit, but not a huge amount in each of those $7k crates...

Another way to tell is to check out Power Equipment prices. Ford and GM make engines and transmissions for use in all sorts of trucks used for towing trains of material in factories and airport baggage and the like. If you can find the prices on this stuff, I bet it is around $10k per engine/tranny, and they are basically what is stuffed into pickup trucks.

Billski

 

[Kyle Boatright]

100,000 aviation engines a year will not happen.
What we need is all new plans built engines. Or partial kits.
Homebuilding is the only solution because labor is deemed free. No liability. No overhead.

Most of us don't have the equipment to self manufacture engine parts. So repurposing off the shelf stuff would the the key, but you'd still need to develop the PSRU, engine controls, cooling system (if liquid cooled), etc. It may still be true that the least expensive practical solution is a used Lycoming or Continental.

 

[BBerson]

Used is always a good choice since they last a long time and never change much.
But I would like to see more kits for the smaller sizes not offered by Lycoming and Continental.
Like a direct drive 100 cu in with many auto or industrial parts such as pistons and valves and custom made crank, case and cylinders and heads. That weighs 90-100 pounds.
Maybe parts could be designed for home manufacture. Obviously a completely different mindset for a designer. Designing for mass production isn't viable.

 

[wsimpso1]

I get when you are designing for a million cars, that the engineering of the production, the associated high speed equipment.... That you get huge economies of scale but at a very high price point to enter the market.
This is also a highly optimized production scenario.
Exclude the drafting aspect.
Think low volume, one unit or five hundred a year. What is the fundamental difference in production costs?

Tim

From having built a dozen here and a dozen there prototypes for early trials and testing, I can tell you that you can plan for less parts. Yep, sand cast all of the aluminum cases, weld up the intake and exhaust manifolds from tube and flat stock, go for low volume ways of making the internals, machine some parts from solid, etc. One path to low costing this product is to use existing pistons and rods, valves and guides and valve trains. And hope the parts keep being made for other purposes so they stay cheap.

Those $1000 transmissions in volumes of 25 to 50 cost on the order of $50k to $100k. Prototype engines are similar. There is no way around it, low volume is really expensive per unit.

Let's just stick with castings. Left and right crankcase halves, sump, accessory case, oil pump housing, intake manifold. North of a million dollars for dies to die-cast them, probably closer to 2 million. Options? Sand cast. Cheapest tool sets to make sand castings only last 50-100 castings, make the walls thicker and the parts heavier. While you are at it, there will be a lot more lost due to voids than with pressure die casting, more rebuilds in-house and warranty to replace cases that crack, and your $25-50k bill for patterns has to be spent again every 50-100 parts. That is $250 to $1000 for each one you cast, not each one you finish. Now if you only make them one at a time, you casting shop is gonna charge you for set up every one you want. First time you have to make and set up the tools, work out your process to get a couple good ones, and then put everything away. A month later you want two more. Well, let's see where we can fit you into production. Next Tuesday, and it will be $1000 to pull the tools, spin up the guys on how to do it, run a couple three to get a good one, and you only want two? OK, $1000 bucks plus the two parts. Nope, run a hundred at a time, and we can waive that $1000 set up charge.

Now let's ship two sets of castings to your machining center, they have to pull the fixtures from storage, idle the work center and install the fixture and cutters set up right for the part, run them. The set up and tear down will take a bunch of time with the cutters not turning that all costs money too. You want to do that in bigger batches than one or two. Somewhere there is a batch size that suits the whole outfit best, between setup costs and costs for parts sitting around waiting to be used.

Then there are gaskets and seals. They almost all have to be custom. Some O-rings and shaft seals might be COTS, but the rest have to be for your stuff. I had prototype O-rings that cost $2500 for a short run tool, $250 for set up and $2.50 a piece for a 1000 piece run. The small ones were well short of a dollar for one of each size in production. Gaskets are fussy thing to get right on production, you want to make a bunch at a time.

Gears with splines and snap ring grooves and mistake proofing features. don't even think about building them one at a time, the set up can be big to get them machined and heat treated right.

You do not want to even talk about building whole engines one at a time. Build all of the pieces in batches of 25 or 50, and you are still in expensive land, but perhaps in a viable zone.

Billski

 

[rv6ejguy]

There are new additive techniques for making casting cores much cheaper than the legacy methods at the quantity levels suitable for this project- dozens per week. Also new additive methods for producing small parts which may be less expensive than traditional CNC methods and in a few more years The Desktop Metal Production System https://www.desktopmetal.com/products/production/ promises to revolutionize 3D metal parts printing. https://www.desktopmetal.com/ https://www.youtube.com/watch?v=EGaUN6ZuHjw

https://www.youtube.com/watch?v=2Kd7rmPrF_w

https://www.youtube.com/watch?v=Z8MaVaqNr3U

There have been some projects run by small teams which have produced limited production engines at unit costs far below what auto OEMs could do the project for. https://www.youtube.com/watch?v=P7ZkRuuU30E https://www.youtube.com/watch?v=4hn8bsGjCik D Motor and Jabiru didn't spend millions getting their designs into production.

I'd never design anything using custom O-rings since OTS stuff goes from tiny to massive in very small increments. Most modern engines use lots of RTV for the last 20 years for sealing many covers, oil pans etc.

Using OTS hard parts is obviously the key to low cost however that will probably never fly with a certified engine or one for LSA which must meet the ASTM specs for QC and parts traceability. Experimental yes.

In short, using some modern technologies, engine development does not need to cost millions of dollars any more with a lean organization of smart people IMO.

 

[BBerson]

That Desktop System said high volume.
Pete Plumb welded up his intake manifold/oil tank from stock tube and plate. Apparently a good way for a homebuilder to make one part instead of traditional casting.
Is anyone doing partial scratch building like that for the complete engine?

The point being to homebuild for satisfaction of doing it rather than buying it.

 

[Hot Wings]

There is no way around it, low volume is really expensive per unit.
i

You have been playing in the 'big boys" sand box for too long.

There are methods to keep costs low for sub 100 unit quantities. Sand casting, for example, is not the only option and I know from experience that simple wooden sand cast tooling with proper gating and such can last far longer than 100 units and produce castings virtually free of voids.

Custom gaskets? Design the thing so it doesn't need anything but OTS gaskets from another high volume production unit - or use the gasket in a can method.

Precision machining? Again design the thing so it's not needed for every part. Design around OTS parts for critical areas, either as is or with minor modifications. We don't even need CNC for low volume production. For some reason everyone seems to have forgotten old fashion JIGs. Simple jigs can be surprisingly accurate and repeatable - even when the part is loaded by unskilled minimum wage labor. The only real downside to using unskilled labor and good jigs and fixtures is that the human QC needed goes up.

But, yes, it does cost a lot more for low unit production. We aren't planning on ever producing Zx10^6 units either so we can make a lot of trades that larger volume companies can't. 2 different worlds.....

 

[Autodidact]

Use as many small block Chevy parts as possible.

 

[tspear]

Use as many small block Chevy parts as possible.

Nah, they have been in production for fifty years. This is old technology, go with something newer please.
Plus, they might stop production of said parts any day now...

Tim (said tongue in cheek)

 

[wsimpso1]

There are new additive techniques for making casting cores much cheaper than the legacy methods at the quantity levels suitable for this project- dozens per week. Also new additive methods for producing small parts which may be less expensive than traditional CNC methods and in a few more years The Desktop Metal Production System https://www.desktopmetal.com/products/production/ promises to revolutionize 3D metal parts printing. https://www.desktopmetal.com/ https://www.youtube.com/watch?v=EGaUN6ZuHjw

There have been some projects run by small teams which have produced limited production engines at unit costs far below what auto OEMs could do the project for. https://www.youtube.com/watch?v=P7ZkRuuU30E https://www.youtube.com/watch?v=4hn8bsGjCik D Motor and Jabiru didn't spend millions getting their designs into production.

I'd never design anything using custom O-rings since OTS stuff goes from tiny to massive in very small increments. Most modern engines use lots of RTV for the last 20 years for sealing many covers, oil pans etc.

Using OTS hard parts is obviously the key to low cost however that will probably never fly with a certified engine or one for LSA which must meet the ASTM specs for QC and parts traceability. Experimental yes.

In short, using some modern technologies, engine development does not need to cost millions of dollars any more with a lean organization of smart people IMO.

Cores by 3D lith might seem reasonable compared to everything else at a few per week. We made soft tools for stampings, lost foam and foam cores, and it made sense when we were making 100 sets, but it was still far from cheap.

There's still lots of places where the stuff inside is big and outside is cramped, and you need enough cross section but can not fit the next standard size. Then if your circumference is an odd size... you end up custom. RTV is nice, we used a bunch of it at Ford and FCA, but it does not go everywhere either. Both are great where they fit and work, and more of it would make sense in airplane power, but in an air cooled head where the temps can go 500 F and the gaskets have to keep sealing, and the exhaust flanges have to stay sealed with the pipe glowing red, well there are still places elastomers won't do.

A low volume 120 hp engines for $12k? I would love to see it and its turboed 160 hp version for $20K but I think I am way more likely to see an EJ257 with 40" Hg and SDS or a turbo-normalized O-320 on the nose of my bird...

Billski

 

[rv6ejguy]

Cores by 3D lith might seem reasonable compared to everything else at a few per week. We made soft tools for stampings, lost foam and foam cores, and it made sense when we were making 100 sets, but it was still far from cheap.

There's still lots of places where the stuff inside is big and outside is cramped, and you need enough cross section but can not fit the next standard size. Then if your circumference is an odd size... you end up custom. RTV is nice, we used a bunch of it at Ford and FCA, but it does not go everywhere either. Both are great where they fit and work, and more of it would make sense in airplane power, but in an air cooled head where the temps can go 500 F and the gaskets have to keep sealing, and the exhaust flanges have to stay sealed with the pipe glowing red, well there are still places elastomers won't do.

A low volume 120 hp engines for $12k? I would love to see it and its turboed 160 hp version for $20K but I think I am way more likely to see an EJ257 with 40" Hg and SDS or a turbo-normalized O-320 on the nose of my bird...

Billski

I think we are talking about less than a dozen per week here, that would be 624 engines per year and I think anyone, including one of the OEMs or clone makers, would be thrilled to sell that many engines. If you make more than that, you can progress to different tooling as you'd have money to pay for it. CNC routered foam cores are cheap to make these days and 3D printing and lith/ sintering processes are evolving almost monthly these days. https://www.additivemanufacturing.media/articles/3d-printing-as-an-alternative-to-patternmaking. New engine manufacture shouldn't be using old tech to produce it in my view.

Silicone O-rings would withstand any temps likely to be found here- VC, accessory case, pushrod tubes, bottom barrel junctions etc. OTS sizes go up in 1/16 increments up to 2 inches and 1/8 increments above that. Easy to design around that. These would not be suitable on exhaust where embossed SS shim types work best.

No, I don't think the discussed engine will be $12K per copy unless it's all machined in China but neither do we need another $25-$30K, 125hp EXP engine. It will be interesting to see what the final price actually is.

 

[Cy V]

Looks like there's an update to this engine on the EPI website. It's just a short paragraph added to the original info:

NOTE: THE ENGINE DESCRIBED HERE WAS DESIGNED AND BUILT BY EPI, Inc. FOR A CLIENT COMPANY AS DESCRIBED BELOW. THAT CLIENT OWNS ALL THE CAD MODELS, DRAWINGS, PROCESSES, COMPONENTS, MATERIALS, TOOLING, AND PRODUCTION RIGHTS AS WELL AS ALL THE INTELLECTUAL PROPERTY REPRESENTED THEREIN. THE CLIENT COMPANY IS ACTIVELY TESTING THE ENGINE EXHAUSTIVELY AND DEVELOPING THE MEANS TO PRODUCE IT IN VOLUME AT LOW COST. THE CLIENT COMPANY IS PLANNING TO ANNOUNCE THE ENGINE TO THE PUBLIC AS SOON AS THEY ARE SATISFIED THEY CAN PRODUCE IT IN VOLUME TO THEIR VERY HIGH QUALITY STANDARDS.

 

[BoKu]

Looks like there's an update to this engine on the EPI website...

Thanks, I'd lost track of this.

Here's the original URL: http://www.epi-eng.com/aircraft_engine_products/aircooled_4-cyl_aircraft_engine.htm

 

[harrisonaero]

In this video it looks like the gentleman at 1:15 is Dr. John Torode (https://vashonaircraft.com/about-vashon.php).

https://www.youtube.com/watch?v=6-HRRGu9pwA&feature=youtu.be

If so that should bode well for the engine and would make an excellent option for their new R7. Though I've never met Mr. Torode I've been involved in aviation all my life and currently as a full time career as a DER. So I have a pretty well established BS sensor for new engines and aircraft designs. I've been impressed with what Mr. Torode has done with Dynon, the hiring of Ken Krueger for aero and design work, and now with leading production of the R7. He seems to be doing it right without a lot of hype and I have a strong hunch he'll sell a lot of planes. Plus he's a physics guy and thinks in First Principles- there's no replacement for that in solid engineering. Just ask Elon Musk.

If this engine succeeds it would be refreshing to give Lycoming, Continental, and Rotax some competition and I wish them the best of luck.

 

[BBerson]

I bet the client is Vashon.
https://www.homebuiltairplanes.com/forums/showthread.php?t=29430&page=5&highlight=New+airplane+at+oshkosh

 

[Cy V]

In this video it looks like the gentleman at 1:15 is Dr. John Torode.

Yeah, I think you're totally right on that.