[llilrex]

I would like to start this thread with an assurance that I am crazy, so with that out of the way I will go on to the more important things.

First I will describe the engine goals,

Extreme reliability, I.E. proven technologies

70 - 100 HP at the crank,

Diesel fuel, as well as JP5 and most other heavy jet fuels

Mechanical injectors, but I may use electronic for sake of performance.

Two cylinders,

Two stroke,

Fully independent injector systems per cylinder,

Non-turbo charged,

Very light weight.

Air cooled,

This engine is to be used for an experimental aircraft.

I am having a problem getting the project started though. So it will be in the design phase for a while. I will keep you posted on the progress but I am going to take it easy for a while, concentrating on the quality of the work.

My next post will detail the rough plans for the design of the engine.

[DSL PWR]

If my memory serves me, aren't most 2 stroke diesel's turbocharged? Other than the obvious reason for a turbo (more air=more fuel=more power) they also allow you to run an ohv system.

So I guess you will be running a conventional 2 stroke (ported cylinger walls). If you are going this way may I suggest you start with a rotax 670 bottom end. These engines have a reputation for taking big power upgrades, and living. I think they were rated at about 120hp stock, and with big bore kits, and NOx would make double that. So at a rating of 70-100hp, (granted there will be more torque than a gas engine) you will have a proven, aluminum 2 stroke bottom end. Then, for the prototype, you will have less to deal with, and that may allow you to focus on the fuel system.

I'd skip the mechanical injection, and go straight to a common rail electronic. This will allow you to tailor the fuel curve in your air cooled engine to keep EGT's under control. Have you also considered a cooling system for the return fuel? I only mention this because on a 100hp aircraft you probably have a small fuel tank, and the whole engine will be running on the hot side, thus heating up your return fuel more than it would be on a 100hp water cooled.

This looks like an interesting project.

[NateTG]

If my memory serves me, aren't most 2 stroke diesel's turbocharged? Other than the obvious reason for a turbo (more air=more fuel=more power) they also allow you to run an ohv system.

Two stroke diesels basically need forced induction in order to scavenge (push the exhaust out of) the cylinders. Typically this means superchargers (compressors driven directly by the engine) rather than turbochargers. Although I haven't seen any examples, it might be possible to use the same sort of crankcase-compressor approach that two-stroke gasoline engines use.

[RICHARD ZASTROW]

llilrex, this may be of interest: http://www.deltahawkengines.com

Dick Z

[llilrex]

Well I am almost done with the design spreadsheet. I am questioning some of the results though. So I will be working on it for a while longer.

Perhaps someone here can answer the questions

If I have a cylinder of .89 grams of air (.000061 slugs) with an injection of .0001337 lbs of fuel equaling 3445J of energy, will in not raise the temperature? With compression of 20.11 the temp will be 986K with 297K air going in, at 1.012 J/g/K this should raise the temp to 4473K. A ratio of 4.5. Now
If the temp is raised by 4.5 times then I would think that the pressure will increase in proportion to (20.11 * 4.5) or 90.46^1.4 * 14.7 for 8060 PSI

This seems high, of course this does not account for the speed of ignition, the burn may last much longer then the conditions outlined here can exist. But still, does this seem right with your experience? Will I have to design a cylinder and head that will withstand these pressures? Or am I way off of the mark here.

[MrWild]

A lot of things are happening all at once. Yes, the point of ignition can reach that temperature, but there are boundary layers on all surfaces that keep that kind of heat at bay. Plus you've got the expansion of the chamber so that cools the burn somewhat. Have you ever poured liquid oxegen on your hand? The boundary layer keeps your hand from freezing as the liquid O2 rolls off and onto the floor. This same affect keeps your motor from melting, and the expanded hot gasses take a large amount of the heat out with the exhaust.

[dynosor]

If I have a cylinder of .89 grams of air (.000061 slugs) with an injection of .0001337 lbs of fuel equaling 3445J of energy, will in not raise the temperature? With compression of 20.11 the temp will be 986K with 297K air going in, at 1.012 J/g/K this should raise the temp to 4473K. A ratio of 4.5. Now
If the temp is raised by 4.5 times then I would think that the pressure will increase in proportion to (20.11 * 4.5) or 90.46^1.4 * 14.7 for 8060 PSI

Are you sure you can burn that much fuel with that much air? If so, the pressure would go up as you calculated if you stopped the crank from turning and the resultant peak pressures would be 4 to 8 times "normal".

Depending on the speed of the engine, diesels are seen to burn fuel at constant volume or constant pressure, or a split between the two.

For a small "high speed diesel" most of the fuel is burnt at near constant pressure because the piston is getting out of the way. This means the peak pressures are much lower at perhaps 2000 psi (at full load) and all your mechanical parts can be lighter.

http://hyperphysics.phy-astr.gsu.edu...mo/diesel.html

http://en.wikipedia.org/wiki/Diesel_cycle

http://www.maritime.org/fleetsub/diesel/chap1.htm

[llilrex]

I saw that the calculations are performed as if it were constant pressure, however it looks to me that they are for the convenience of calculation and visualization, under ideal conditions. However I no experience in this assumption and that is why I posted the math. In my mind I am considering worse case design scenario. Where the crank is stalled at TDC at of course maximum pressure. I am aware that this is an unlikely event and it is probably unrealistic to assume the condition. As for the constant pressure PV chart I wonder if in fact it really looks like that, for instance the crank is at TDC moving at x velocity, the burn starts but progresses faster then the crank can get out of the way, is it not reasonable to assume that under real world conditions where the piston can not move out of the way due to kinematical restraint that the pressures would be seen as it would necessarily rise in lue of a volume increase.

but all that said I am still wondering what the curve would look like and what engineers base the strength calculations on.

I am still working on the sheet as I have found a very off calculation concerning the kinematical interpretation of the power produced. it said that due to average torque on the crank I was producing over 450hp which is simply not possible.

Any thoughts on this would be great.
To summarize the questions asked:
What does the real load look like? Is it really constant pressure as the PV chart indicates? What do engineers typically base the strength calculations on?

Thank you for your comments!

[dynosor]

You are right that combustion pressure curve are not geometric like the ones shown in the above link. Those are shaped that way for simplification and calculation convenience.

I found this presentation on diesel combustion with theoretical and measured pressure/volume graphs, and power calculations:
http://www.sfws.auburn.edu/plm/Web%20Engines.ppt

and spark ignition:
http://www.tfxengine.com/NitroEngineData.html

http://www.tfxengine.com/NaturallyAs...ngineData.html


I think designing your engine for "locked crank" combusition pressure may be one way to add a factor of safety, but it is a little unconventional.

Knowing what the real peak stress is going to be and then over designing for that would usually be based on a list of factors, such as:

Fatigue: If loads vary significantly, suddenly, continuously and especially if the load reverses (more than applied and released), then increase calculated strength by a factor of 10 based on material yield strength.

Parts like connecting rods must be kept light or they increase loads on other parts. Instead, measures are usually taken to improve conrod fatigue strength such as bead blasting (surface compression) of the part's outer "skin"; and super finishing the bores of piston pins. In other words, reduce stress magnifiers.

Heavy conrods and pistons drive the need for a heavy crank and larger bearings, and reduce the potential operating speed of the engine.

If a part is subject to corrosion, double the design strength again, or schedule service intervals to replace parts before they fail.

Design the crank shaft not only to have sufficient strength, but to achieve a natural torsional frequency that is outside the operating speed range. Or, make the parts lighter, but do not allow the engine to operate at speeds that excite fatigue inducing vibrations - it was common to run ship and aircraft engines up past "bad" speeds before loading them appropriate prop pitch - see link on old submarines above.

For reliability, you end up with a heavy, less efficient engine or vehicle, or a lighter engine with higher cost super finished and balanced components, and/or shorter parts replacement service intervals. For instance, aircraft engines must be light to fly, but are not allowed to fail in flight.

[llilrex]

does anyone know a good injection duration?

[dynosor]

does anyone know a good injection duration?

Fuel may be injected in one to three distinct pulses, depending on load. At high load, the total injection period can span 45 degrees of crankshaft rotation.

For an engine running at 4000 RPM (67 revs/second), total injection period is 0.0019 second at high load. At low load, a single injection pulse would be shorter than 0.001 second.

You need to consider piston velocity. For a diesel car engine with a bore and stroke of about 3.5", peak power speed is typically 4000 RPM with max RPM of about 4500. For truck engines and marine diesels you notice that the operating RPM comes down as stroke length increases to hold mean piston velocity near constant.

You need to determine what piston velocity to aim for and match that with your smaller engine. If your engine's stroke will be 1.75" then max power revs can approach twice 4000 RPM. The problem is that when you double RPM, inertial loads are 4 times greater...

Injection periods probably do not vary by more than 2:1 for all types of diesel engines under full load conditions.

[llilrex]

thanks

[dynosor]

I forgot to emphasize that the one or two thousands of a second injection period only apply at 4000 RPM in that example. At 2000 RPM you would have twice the time to inject the same amount of fuel per cycle.

As it is, rather than inject fuel continuously at a lower flow rate over a 2/1000 second period, injection is broken into multiple distinct bursts. In other words, at lower power demand the injection period is shorter, but occurs with the same intensity to ensure fuel oil is broken into small droplets.

More on injection:
http://www.testing-expo.com/usa/06co..._macdonald.pdf

http://www.erc.wisc.edu/modeling/mul...Das_Delphi.pdf

[llilrex]

I just bought a book detailing procedures for calculating forces on the crank and other issues that I was having. I hope this project plugs along, hate to leave another one on the shelf.

Thanks for the reply; it gives me a better idea on the particulars of these engines.

[dynosor]

What is the name of the book, if you please?

[llilrex]

Diesel Engine Engineering : Thermodynamics, Dynamics, Design, and Control, Vol. 143
by Andrei Makartchouk, Makartchouk Makartchouk

Publisher: CRC Press
Pub. Date: March 2002
ISBN-13: 9780824707026
Sales Rank: 216,517
392pp

http://search.barnesandnoble.com/Die...4707026/?itm=8

[dynosor]

Diesel Engine Engineering : Thermodynamics, Dynamics, Design, and Control, Vol. 143
by Andrei Makartchouk, Makartchouk Makartchouk

Publisher: CRC Press
Pub. Date: March 2002
ISBN-13: 9780824707026
Sales Rank: 216,517
392pp

http://search.barnesandnoble.com/Die...4707026/?itm=8

Thanks. This looks like a really useful book. I ordered one too despite the hefty price.

[llilrex]

Yea it was pricy but it proved quite useful, I am still trying to decipher the math though. It is just a matter of wrapping my head around it. I am still trying to find out how my calculations are wrong, but the graphs that I am coming up with look quite similar to the ones shown in the book. I can’t wait to make some chips on the project!

[llilrex]

Does anyone have detailed drawings or pictures of the crank shaft for a Rolls-Royce Merlin XX or similar engine?

I am curios on how the oil passages and the engineering was done on this part. I have seen some pictures and a partial cutaway of the crank but I need some more to go off of. Thanks.

[dynosor]

Does anyone have detailed drawings or pictures of the crank shaft for a Rolls-Royce Merlin XX or similar engine?

This seems to be it for a few Bob: http://www.aviationshoppe.com/catalo...rlin-p-35.html