Technical Articles and Product Descriptions

Reference Materials

Additional Products

Race Engine Technology

When you read this page, keep in mind the fact that one of the primary specialties of EPI, Inc. is Aircraft Engine Conversions, and two of our primary product lines are Propeller Reduction Gearboxes and Liquid-Cooled V8 Aircraft Engines.

Yes, we like doing engine conversions. Yes, we like to sell engine and gearbox systems.

BUT, when you apply the cold, hard engineering realities, it turns out that most of the time, replacing a Lycoming or Continental air-cooled aircraft engine with a liquid-cooled piston engine is a bad idea.

That being said, there are a few cases where it is an excellent choice. If you are convinced that you have decided on a sufficiently-reliable powerplant to replace the existing one, here are some thoughts to help decide whether to go further.

1. When you do a true, accurate, apples-to-apples comparison of the total, firewall-forward weight of a complete, flying Lycoming or Continental powerplant against that of a complete, flyable liquid-cooled powerplant of the same or greater power (real, measured SAE Horsepower, not "BlantonPower"), the liquid-cooled installation will almost always outweigh the Lyco-Nental by a considerable amount. There is a table at the end of this page showing details supporting this statement.
2. NEXT, the CG of a V8-Gearbox powerplant will typically be quite a bit farther forward than the CG of the air-cooled powerplant.
3. The combination of the greater weight and greater overhung moment means that the engine mount structure of a V8-gearbox powerplant will typically be heavier than that for the air-cooled engine you are replacing, because it has to support (a) a larger weight (b) whose CG further from the firewall. (There is a short discussion of mount structures later on this page, and a complete page dedicated to engine mount design.)
4. The combination of greater weight and greater overhung moment applies proportionately larger tensile, compressive and shear forces to the aircraft engine support structure behind the firewall. MOST airframes (fuselage / nacelle structures) are not strong enough to support these larger forces. In fact, in most conversions we have done or assisted with, significant airframe reinforcement / redesign was required.
5. The greater weight of the powerplant and the structural additions to the airframe to support it will increase the empty weight of the aircraft, which will either (a) reduce the payload (if you understand dynamic airframe loading) or significantly reduce the ability of your aircraft to survive in-flight maneuvering loads (if you DON'T understand dynamic airframe loading and arbitrarily decide to "declare an increase of the gross weight").
6. The combination of greater powerplant weight and further-forward powerplant CG location will move the empty CG of the aircraft further forward. You will have to compensate by moving the wing forward or by moving some other masses rearward to re-establish the proper relationship between the aircraft CG and the wing center of lift.
7. Unless you move the wing forward (not usually a simple task, nor one without other potential pitfalls) the adjustments needed to bring the aircraft empty CG back into range, combined with the more-forward CG of the powerplant, will increase the Mass Moment Of Inertia (mmoi) of the aircraft about the pitch and yaw axes (AND, in the case of a twin, around the roll axis as well). That change alone can have a dramatically undesirable effect on the flying qualities of a previously pleasant-flying aircraft. It reduces the stabilizing influences of the rudder and horizontal stabilizer on the aircraft. It makes the aircraft less responsive to control inputs, (more "sluggish"). It will reduce the resonant frequencies of the aircraft about the pitch and yaw axes (and roll in the case of a twin). That change in resonant frequencies can introduce some nasty dynamic behavior in flight, such as pitch or yaw instability.

The Bottom Line is: JUST HOW COMPETENT AN AIRCRAFT DESIGNER ARE YOU?

A Very Bad Example

One example of a really bad conversion idea is the replacement of the IO-540 recommended for Van's RV-10 aircraft with an LS-1-based aluminum V8 engine. This glaring example of "denial-of-reality-engineering" was actually written up in one of the goofy experimental magazines as being the greatest thing since sliced bread. This idea exhibits ALL of the problems discussed above.

And the bad ideas get even WORSE.........Here is an example of some of the frequent inquiries we get regarding conversions:

Dear EPI
I am looking for an engine to fit my RV7A, we have an LS1 out of an 2001 Firebird with about 65k miles on it, how much will it cost to have your reduction unit intalled and have the LS1 engine rebuild to your specs, I would also need the engine mount and electrical wiring done with a new computer, also a veriable (sic) pitch prop governor.
thanks

xxx xxxxxxxxx

And here is our standard response to such inquiries:

Mr. xxxxxxxxx:

Thank you for your inquiry about our products.

Regarding your proposed engine project, it is a strict company policy that we will neither sell any equipment nor do any work at all on engine conversions which obviously violate many of the known laws of physics and engineering.

PLEASE read this page on the EPI website:

www.epi-eng.com/aircraft_engine_conversions/additional_weight_considerations.htm

It might also be helpful if you read carefully the entire section on the EPI website regarding engine conversions:

www.epi-eng.com/aircraft_engine_conversions/conversions_contents.htm

Sincerely,

Jack Kane, EPI, Inc.

An Extremely Bad Example.

We worked on one engine conversion project on a popular piston twin where the goal was to replace the aged and overtaxed Continental GTSIO-520 engines with a pair of certified, liquid-cooled V8 engines which had substantially greater takeoff and cruise power.

The takeoff, cruise and engine-out performance of the V8-powered, Experimental-R&D-certified prototype aircraft was quite impressive. However, the installed powerplant weight was a flaw which doomed the project..

The total, firewall-forward weight of the original GTSIO-520 ( engine, gearbox, prop, turbo, heat-exchangers, mount, cowling, baffles, ductwork and plumbing ) was about 930 pounds. The comparable firewall-forward weight of the V8 installation (absolutely apples-to-apples comparison) was slightly over 1260 pounds, which added about 660 lb. to the empty weight of the aircraft.

That increase in the aircraft empty weight had a number of undesirable side-effects, including:

1. The wing-structures to which the V8 engine mounts attached were unable to safely support the new loads;
2. The large increase in nacelle overhung moment reduced the resonant frequencies of the wings in torsional and flapping modes;      and, BTW,
3. The resulting aircraft was a functionally-useless "full-fuel, zero-person" aircraft.

The project manager declared the weight problem to be of no consequence, because, he claimed, he could easily obtain, as part of the STC, a gross-weight increase based on the claimed "similarity" of this model of the aircraft to a turbine-powered model which has a certified gross weight approximately 1000 lb. greater, and which has substantially lighter engines.

An unfortunate snag to that plan was the fact that the allegedly "similar" aircraft was very DISSIMILAR in the wing structure to which the engine mount attaches. And there were other serious snags, including the ones mentioned in (a) and (b) above.

COLD HARD NUMBERS

Admittedly, that was an extreme example. But look at the following table, which shows representative comparisons of the installation weights of several Lycoming engines against the lightweight (for a Liquid-Cooled-V8), normally-aspirated, 505-HP EPI Gen-1 Aircraft Engine. The weight of a powerplant based on the GM Gen-4 engines ( LS-2 / LS-7 ) on an apples-to-apples comparison, will weigh about 25 pounds less. 

The weights listed in the table are generally from measured components being used on actual aircraft. The Lycoming weights are direct from Lycoming documentation, verified from hands-on experience. AND, this is not an exhaustive list. It shows only the major items.

PLEASE NOTE: The Gen-1 V8 in the chart below DOES NOT include any turbocharging considerations. To estimate a turbocharged V8, add 60-90 pounds to account for the weight of (a) the turbocharger and wastegate, (b) turbo support structure, (c) intercooler, (d) inlet plumbing and miscellaneous other stuff.

These numbers can be made to vary a bit by altering the specific configuration. Composite propellers can be substituted for metal ones.  Cowling, ductwork, and engine mount numbers can vary.  Things can be omitted. But the table is a good representation of reality, taken from actual flying aircraft.

Note that the only engine in the table which makes sense to replace with a V8 is the Lycoming TIGO-541 (the engines on a Piper Navajo). There are a few others which make sense too, including Pratt & Whitney radials (R-985 and R-1340) and most installations using the GTSIO-520 / 550. But we have had people come to us wanting to replace all the other engines listed in the table with some form of liquid-cooled V8.

The engine mount numbers in the table above are quite different. Why? Lycoming engines are cantilevered off the back face, which allows the mount to be quite short and stiff, therefore light. Here's a 19-pound mount for a 300 HP Lycoming IO-540 engine (which includes the nosegear structure).

Here is the mount (still in the fixture) for one of our cropduster conversions. The mount weighs 53 pounds, and supports all the FAA combined loads from a 750 HP, 1235-pound V8 powerplant with a comfortable margin. (The little hose at the front is for replacing the air and moisture inside the tubes with argon.)