Injected Diesel 56cc 2 Stroke, Will it ever work?"

[Lloyd-ss]

From the looks of your cylinder liner, you plan to have cooling?

Forgive me if it was mentioned but are you fitting a coolant pump?

Will you eventually add a demo load like a generator to the engine, or will it be a very large tabletop proof of concept engine?

Great questions!
I have allowed for water at the top end of the liner, and also in the head but have not done the drilling in the head yet. Luckily the stainless that the head is made out of drills nicely.
Something I do when working on design-build projects is try to not box myself into a corner in the early stages. I never know what kinds of things I am forgetting about, but I still try to leave some space for them. It often works out.

Lloyd

 

[Lloyd-ss]

From the looks of your cylinder liner, you plan to have cooling?

Forgive me if it was mentioned but are you fitting a coolant pump?

Will you eventually add a demo load like a generator to the engine, or will it be a very large tabletop proof of concept engine?

I forgot to answer the part about the load for the engine.
Yes, I will definitely have a load for running and testing it. That is PART 2 of the engine building project FUN!!
If you look at my you tube channel you will find a very eclectic mix of videos, but you will definitely see lots of testing "to failure" and experimentation. There will be no testing to failure this time ( I hope), but I definitely want to try it with the dyno instrumentation I have. I have been looking a strap band brakes of 8" diameter to use as a load around the flywheel.
That will be a lot of fun, and I plan on being cautious so that I don't too many "points of failure."
Lloyd

 

[Lloyd-ss]

COMPRESSION RATIO, TEMP, PRESSURE, ISOTHERMAL, ADIABATIC, and GAMMA
TESTING
@Nerd1000 already has the "nerd" name claimed, and he has proven to be worthy of that title, but it is my turn to throw in some nerd/geek stuff myself. I love spreadsheets and calculations and do a lot of that behind closed doors to avoid embarrassment. Thermodynamics falls into that weird category too. "So, I hear that you actually understand AND LIKE thermodynamics....... I'll be seeing you. Bye." But I neither understand nor like thermo. To me, it is a subject where for centuries, very smart people have been trying to quantify with all sorts of rules and formulas how to explain the fickleness of the subject. You have to guess at what secret formula to use to solve each different problem. It is fun when you choose the right path (method), but the wrong path will lead you straight to hell.
Compression ratio is easy as is temp and pressure so long as you make the correct original guesses.
Is it Adiabatic where all the heat of compression stays in the trapped air (check, the correct answer) or is it isothermal, all of the heat of compression is transferred into the metal of the engine (wrong, put your head on your desk). Or is it somewhere in between, uhhhh, damn, that might be correct, too. See what I mean about being fickle?
I won't bore you too much (I've already done that but if you are still with me, thank you!
Diesel compression is basically adiabatic. You have a starting and ending pressure (P1 and P2), a starting and ending temp (T1 and T2), and a starting and ending volume (V1 and V2). And most importantly, the adiabatic index, gamma, of air, which is 1.4 .Other gasses and gas mixtures have other values of gamma.
In the following calcs, there is a spot for an exponent (I call it the fudge factor) where the full value of gamma 1.4 is used for adiabatic processes, and at the other end of the spectrum, the value 1.0 is used instead of 1.4 for isothermal processes. And if that doesn't give you the right answer, you guess at a number between 1 and 1.4. Please pardon my poetic liberties and sarcasm.
Now for the real answers with numbers for adiabatic diesel:
We Know P1,V1,V2, and T1. We want to solve for P2 and T2.
P2=(P1V1^gamma)/(V2^gamma)
(use gamma of 1.4,but you can go all the way down to using 1.0 for isothermal)
T2=(T1*(V1^(gamma -1)))/(V2^(gamma-1))
(same footnote of using 1.4 all the way down to 1.0)
P1= 1 bar(14.7psi), V1=56cc, V2=2.8cc(20 to 1 ratio), T1=22C(72F), gamma=1.4
Answer P2= 66.3Bar (974psi), T2= 706C (1303F)
But now, if we say that some of the heat of compression is absorbed by the metal of the engine and use a value of 1.2 instead of the full value of gamma of 1.4, we get:
P1= 1 bar(14.7psi), V1=56cc, V2=2.8cc(20 to 1 ratio), T1=22C(72F), gamma=1.4 use 1.2
Answer P2= 36.4Bar (535psi), T2= 265C (508F)

Ain't that something. See how an ice cold engine can suck all the power out of a diesel.

I am relatively sure everything I am saying, and my calculations, are correct, but feel free to give constructive criticism if you like.
Lloyd

 

[Lloyd-ss]

Leaky compression Test Data

I attempted a prelim compression and temperature test yesterday to see if the test equipment was functioning. The test setup was ok, but the compression in the engine is not yet up to par, and the drill that I am driving the engine with is not up to the task. It only got up to 8.3 bar (120psi) and 65C (149f) but it acted as an indicator. I will re-test after fixing the problems on the engine.

This pic is of the plug that fits into the injector hole. It is milled out to receive the thermocouple which is held in place and sealed by 2 o-rings.



Installed in the head. The washer and 2 screws keep it from popping out.




View with the steel tube and T-couple wire going to the gauges.



The compression tester and T-couple display. Both measurement tools pulsate with the compression stroke, but do stabilize after several cycles.

 

[ninefinger]

Do you have a check valve inline anywhere on the pressure measurement (ideally inside the plug)? Curious as the length of tubing may be significant enough to impact the CR and the max pressure achieved.

 

[Lloyd-ss]

Do you have a check valve inline anywhere on the pressure measurement (ideally inside the plug)? Curious as the length of tubing may be significant enough to impact the CR and the max pressure achieved.

Ahhh, good point. There is about 8" of 1/8" O.D. steel tube coming from the head, then an adapter with a several cc's of volume, then a Schrader valve as a check, then 16" of rubber air hose, then some standard air fittings adapted to the gauge and then a check and pressure relief button at the gauge. So, yes, lots of extra spongy air volume. I can eliminate the rubber hose and other dead volume and verify the correct location for the check to get more responsive readings. Thank you!

 

[Nerd1000]

Do you have a check valve inline anywhere on the pressure measurement (ideally inside the plug)? Curious as the length of tubing may be significant enough to impact the CR and the max pressure achieved.

That's significant even for somewhat bigger engines. When my partner and I bought our sailing boat it came with a Mercury 3.3 HP outboard, about 80cc if I recall correctly. It compression tested way too low, but you could see the cylinder was in good condition. When I used a tester with a check valve it tested perfectly normal.

 

[Roger B]

From the looks of your cylinder liner, you plan to have cooling?

Forgive me if it was mentioned but are you fitting a coolant pump?

Will you eventually add a demo load like a generator to the engine, or will it be a very large tabletop proof of concept engine?

I guess this is referring to my cylinder liner. Yes, it is designed for water cooling and will have a water pump to allow load testing if it ever runs well enough. The engine build is here:

https://www.modelenginemaker.com/index.php/topic,8690.0.html

I load test all my engines when they are running well enough. The load is a cheap DC motor connected to a resistor load bank. It does not give an accurate output measurement for the engine but allows comparisons between different settings and components. Someone on here (possibly Nerd ???) was developing a proper dynamometer.

These two engines have pumped water cooling and go through various load tests:

https://www.modelenginemaker.com/index.php/topic,7883.0.html

https://www.modelenginemaker.com/index.php/topic,7701.0.html

I also did a ‘real world’ test with the fuel injected 12cc twin. Here it is powering an electric car I built for my children many years ago. The motor I am using for a generator didn’t really have the correct characteristics so it was like driving in too higher gear, but it moved

 

[Roger B]

Those laps look high quality like they would control roundness, straightness, all that good stuff. I don't know if you mentioned in detail how the diameter adjustment works and how you load the lap with grit. That video really emphasizes the difference between quality single-purpose built tools and having to make-do with what is on hand.

The Acrolaps are quite simple, there is a cone shaped piece that screws into one end of the brass bush to expand it, the other end of the bush is screwed onto the mandrel. The expander and mandrel cover a range of sizes, in this case 15 – 25 mm which covers my range of engines.

To start lapping I make sure that it is set a little to small and squirt 2 or three ~1cm long stripes of diamond paste on different parts of the lap plus a few drops of light machine oil. I then slide the cylinder over the lap to distribute the paste and oil. Next I expand the lap until I feel a little resistance to movement. At this point I start the lathe and move the cylinder back and forth until the resistance drops. Then I will stop the lathe and screw the expander in a little bit more until there is some resistance again and continue lapping. At some stops I will add some more oil. Finally when I think all the machining marks have gone I will stop and clean the cylinder in an ultrasonic bath too remove the traces of diamond paste. There is a lot of hands on experience in this, how much paste and oil to use, how much to expand the lap, what speed to use (I tend to use 250 or 500 rpm) and how long to keep lapping at each step.

They are a reasonably priced system for hobby/light industrial use, a 20mm set is around 25 USD. The Swiss precision version is around CHF 500 for a 20mm lap, mandrel and expander.

 

[Roger B]

From the looks of your cylinder liner, you plan to have cooling?

Forgive me if it was mentioned but are you fitting a coolant pump?

Will you eventually add a demo load like a generator to the engine, or will it be a very large tabletop proof of concept engine?

I guess this is referring to my cylinder liner. Yes, it is designed for water cooling and will have a water pump to allow load testing if it ever runs well enough. The engine build is here:

https://www.modelenginemaker.com/index.php/topic,8690.0.html

I load test all my engines when they are running well enough. The load is a cheap DC motor connected to a resistor load bank. It does not give an accurate output measurement for the engine but allows comparisons between different settings and components. Someone on here (possibly Nerd ???) was developing a proper dynamometer.

These two engines have pumped water cooling and go through various load tests:

https://www.modelenginemaker.com/index.php/topic,7883.0.html

https://www.modelenginemaker.com/index.php/topic,7701.0.html

I also did a ‘real world’ test with the fuel injected 12cc twin. Here it is powering an electric car I built for my children many years ago. The motor I am using for a generator didn’t really have the correct characteristics so it was like driving in too higher gear, but it moved

 

[Roger B]

Ahhh, good point. There is about 8" of 1/8" O.D. steel tube coming from the head, then an adapter with a several cc's of volume, then a Schrader valve as a check, then 16" of rubber air hose, then some standard air fittings adapted to the gauge and then a check and pressure relief button at the gauge. So, yes, lots of extra spongy air volume. I can eliminate the rubber hose and other dead volume and verify the correct location for the check to get more responsive readings. Thank you!

This is my compression tester, I have tried to keep the non return valve as close as possible to where the injector nozzle would be. I have had problems getting the ball to seal properly at the end of a deep hole.

 

[Lloyd-ss]

snip.................. Someone on here (possibly Nerd ???) was developing a proper dynamometer.
.................................

Hi Roger,
Actually, that was me who built the prony brake dyno with data logging capability. It utilized a load cell, a reaction arm, and a hall sensor, and an arduino micro, and recorded the data in an excel spreadsheet.

Here is the build thread and demo for the dyno.
https://www.homemodelenginemachinis...d-with-data-logging-real-cheap-success.35924/

And here is a screenshot of the data output from an electric motor I tested.

And the complete dyno set-up

Lloyd

 

[Lloyd-ss]

The Acrolaps are quite simple, there is a cone shaped piece that screws into one end of the brass bush to expand it, the other end of the bush is screwed onto the mandrel. The expander and mandrel cover a range of sizes, in this case 15 – 25 mm which covers my range of engines.

snip....................

Roger, thank you for the lap info. That looks like a nice system.
Lloyd

 

[Lloyd-ss]

I guess this is referring to my cylinder liner. Yes, it is designed for water cooling and will have a water pump to allow load testing if it ever runs well enough. The engine build is here:
snip...........

Hi Roger,
Thanks for that info, but I do believe he was referring to the build of my 56cc diesel (this topic) and I answered that question in post #281.

Here is more info about the water jacket arrangement for the diesel.

Here is a view looking into the bottom of the outer jacket, with the liner inside.



Here is the liner with one of the two o-ring separators that will isolate the water jacket from the pressurized air system. The scavenging ports will be fully isolated, as will the water jacket.

Lloyd

 

[Lloyd-ss]

This is my compression tester, I have tried to keep the non return valve as close as possible to where the injector nozzle would be. I have had problems getting the ball to seal properly at the end of a deep hole. snip........
View attachment 158238

Hi Roger,
I agree that leaks can be a real source of frustration.

Here is the type of check valve design that I use whenever I can. It is commonly used in high pressure airgun fittings and will work from 2 bar all the way up to 300 bar. The simpler check at the top is my favorite. It is very compact, needing a little bit of length but very little diameter. It is built with almost no clearance around the O.D., but the air will leak by in the correct direction only. It requires a hard o-ring, 90 duro or so. It would normally fit inside of the male 1/4" quick disconnect fitting at the left. If the length is made such that the tip of the check sticks out the end of the fitting, it can then be tapped on a hard surface to let the pressure off, usually with a loud pop. If the o-ring is too soft, they can end up accepting air, but not releasing it. The check shown would probably fit inside the fitting you have drawn. When everything works well on the check, it will work for years.
Lloyd

 

[Roger B]

Thank you I will look into that. I have been using a ss ball in a lapped seating as I do for my injection pumps which seem to work well in an oily environment. As the compression tester is dealing with hot air another solution maybe required. I am interested in your temperature measurements. As I see it the peak temperature is only present for a short time and the thermal inertia of the thermocouple may be to much.

 

[Lloyd-ss]

Hi @Roger B
Getting instantaneous pressure &temp readings is difficult. But the both have a period of several revolutions to build and stabilize in this situation. In the past I have read the values directly on an oscope to avoid the processing lag for the standard display. It might be difficult to see, but at the current time, the thermocouple junction is touching the aluminum hole adapter which doesn't help the situation. I will try to get the junction into free air and see what that does. These are fun diversions.
Lloyd

 

[Roger B]

Those are some interesting thoughts I got out my compression tester, a small thermocouple and the smallest O rings I have, 2mm bore 1mm section. I think I could open the nozzle out to 4mm bore to take an O ring however I don’t think I could get the thermocouple in as well. I might make a plastic nozzle and glue the thermocouple in to it. This should reduce the heat loss. I can then measure the pressure in one test and the temperature in another using the same speed and time.

This is the way my compression tester operated. There is probably too much dead volume after the non-return valve and the pressure will also drop as the air temperature drops. There could also be a leak somewhere else.

 

[Nerd1000]

I also did a ‘real world’ test with the fuel injected 12cc twin. Here it is powering an electric car I built for my children many years ago. The motor I am using for a generator didn’t really have the correct characteristics so it was like driving in too higher gear, but it moved

Petrol fuel injected? Do you have a thread on this one somewhere? I'm curious how you decided to do the fuel metering for that, I know there are many possible ways and have sometimes idly thought about how I might do it.

If I recall what I read correctly, the WW2 German engines with direct injection used some sort of 3-dimensional cam to control the fuel rack position based on manifold pressure and engine speed. It seems like that would be difficult to make in a home shop.

 

[Roger B]

I have two engines running with petrol injection. The full details are on MEM. Both use manifold fuel injection with the same design of mushroom/poppet injector I use on the diesel and have slide throttles.

The 25cc single has a variable stroke pump where the speed - leakage characteristic seems to match well and it will run over the full speed range without adjustment. The injection is timed during the inlet stroke and the pump is operated by an excentric to give an extended injection period. This engine has an electric fuel pump to keep a pressure around 0.8 -1 bar at the injection pump inlet to stop vaporization.

The twin has helix controlled pump driven by the crank shaft to give an injection per rev to match the induction times with a 180° crankshaft. This has a diaphragm pump to maintain the fuel pressure at the injection pump inlet. The pump characteristics don’t match so well on this engine and it has a cam connected to the throttle to reduce the pump working stroke as the throttle is opened and the speed increases, reducing the internal leakage of the pump. The cam was designed by measuring the rack position required for best running at various throttle openings and loads.

This achieves my desire for flexibility, I can put the engine on and off load without it stalling and can rev up under load. Most ‘simple’ model engine carbs cannot do this, they are designed to drive propellers which are a very benign load. I have also run this engine with a constant depression (SU type) carb, the results of out put and flexibility were very similar to the fuel injection.