Diesel engine performance testing question 1

[braddles90]

Hi all,

Sorry if this is too simple a question, but I'm wondering whether I'm missing something thinking through a problem at work.

We currently have a 6 cylinder CAT diesel engine (mechanical fuel injection, turbocharged) on a dyno doing some performance testing (used on mobile machinery), and we've found that when developing the torque curve for the engine, if we start at high idle, load the engine, then check the maximum torque output at 100 rpm increments in decreasing order, we get a set of numbers (say we got 400Nm @ 1900rpm). We then drop the engine to 1900rpm, no load, increase the load, and we find we can only reach about, say, 325 Nm until the engine speed starts to drop off.

What would be the main reason for this? My initial (simple) thought was that it should reach the 400Nm as measured in the first test, however after discussing with the testing guys their argument was that just increasing the load at 1900rpm won't let the engine develop enough boost to reach its full power, but when you load it up at rated power then start generating the curve backwards, you already have the speed in the turbo to max out the torque at lower RPM.

It still seems a little odd to me - maybe I'm thinking about it too simplistically, and the above argument does make sense - but I'm wondering if there is anything else at play? Or is this a sign of a poorly matched turbo?

Any thoughts? Thanks for the help!

 

[dgallup]

The turbo should be able to come up to speed, it might take a little while but only a minute or so. It sounds to me like there may be a lot of hysteresis in the governor and it is not getting to full fuel delivery. A little hysteresis is normal but 75 Nm is a lot. How old is this engine? The pump kit may need to be rebuilt. Sometimes the fuel delivery cam gets a groove worn in it and the follower will catch and hold until the flyweight loads build up enough to ride over it.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[ivymike]

That's indicative of some sort of malfunction if you're spending enough time at each point that dyno inertia isn't a factor.

The torque curve should be more or less the same up as it is down.

If the turbo is the problem as suggested by your colleagues then I think the rack would still go to the same point and you'd see smoke due to lower-than-desired AFR. You probably wouldn't lose so much torque either.

Maybe it's a governor issue, is it possible that there's a control or measurement issue with the dyno?

 

[SomptingGuy]

If this were a gas turbine engine (as it sort of is really), I'd find the observed behaviour more easy to believe (those things are mad). Maybe the speed/load map really does have spots that can't be crossed directly and have to be driven around?

- Steve

 

[ivymike]

hey, any reason you're not using a published torque curve from the manufacturer? (vs trying to dyno it yourself)

 

[catserveng]

If you can provide a model and serial number I may be able to get you a better answer. In general when dyno testing an engine, as ivymike points out above, it should be pretty close up and down. There is some hysteresis, mainly due to the condition of the mechanical governor. Are you taking dynamic fuel rack measurements when doing the testing?

I used to run a dyno and test stand at a dealership many years ago, would likely suspect an issue with the hydramechanical governor before the turbo. I have run into a similar problem on engine governor equipped with a smoke limiter that was improperly adjusted, but again, need details as to which engine and fuel system you have for specifics, CAT made a LOT of 6 cylinder engines for off hiway equipment.

MikeL.

 

[braddles90]

Cheers guys, thanks for the replies.

A bit more info - I work for a mining machinery OEM where we run with pretty tight surface temperature restrictions. As such we do crazy things like water jacket turbo compressors, exhaust piping, add water baths to cool the exhaust; after all these mods we need to do our own performance testing to check the impact of the changes we design. We've used this engine in the field for 10+ years, but we're currently doing some other testing and the test facility wanted to check the engine performance against our type testing and found this behaviour. It's an essentially new engine (maybe between 500-1000 hours runtime), and has only been used as a dyno rig.

I haven't got the exact numbers on hand from the test I saw, but there was a definite ~100Nm drop when we tested it on the way up than on the way down at rated torque speed.

I know we have our own arrangements with CAT for our fuel rack settings (I believe that we've derated that model from the standard setting to help drop surface temperatures), but interesting that it could be an improperly adjusted smoke limiter/governor. I think the guys that built the engine had been playing with the fuel settings because this test engine had been supplied standard (not to our requirement) - that could be something we should check again.

ivymike - I thought that could be the case, but we don't really see any extra smoke, and wouldn't the governor limit that somewhat because it wasn't seeing enough boost to increase the fuelling any more?

catserveng - It's a 3126 engine, I believe we use the marine variant from CAT then modify a few things. We haven't been monitoring the fuel rack, only the throttle position (I'll see if we can get the guys to monitor that if they do any more testing). I'll see if I can get a serial number off the guys as well.

Thanks!

 

[ivymike]

I thought all the marine 3126s had higher horsepower than what you've measured?

There were some 3126s with air-fuel ratio control, and some without ... on the ones with AFRC you could conceivably have fueling cut back due to low boost, whereas on the non-AFRC you would not. Is there an air line to your governor?

There were some 3126 engines with "dual horsepower" governors on them, maybe you've got a malfunctioning one of those? The dual horsepower mechanism wss used to limit horsepower under certain operating conditions, and it was controlled via a solenoid on the governor.

There were also some 3126 ag engines with a torque limiter feature. The torque limiter is used in conjunction with a torque cam in order to control torque rise on the engine, to give a torque rise with two slopes. In these ag applications, there is a rapid torque rise from rated torque to peak torque. At lower speeds, the rise is more gradual.

 

[braddles90]

Hey all,

Sorry for the radio silence, that misbehaving engine got taken off the dyno and I never got around to digging up exactly what the governor style is on our engine, so sorry about that.

We've got a different engine on the dyno at the moment though, another mechanical engine but with a different rotary-style Delphi fuel pump - we checked the engine torque both up and down the curve and it made it both ways easily, so we're pretty convinced the guys who set the 3126 up haven't set the A/F ratio controller properly. Not sure when that engine's being dragged out again though, but that lesson helped me check this new engine is running properly.

Ive, just to answer you questions - I threw some general numbers out there instead of the actual ones (I'd forgotten the actual measurements when I made the power, but you're right, the numbers were more like peak torque of 750/800Nm, and we could only get it to 650 Nm on the way up. My mistake!). I think the AFRC controller is definitely the one we have, we have the air signal line to the pump hooked up, and I'm sure at some point that kind of funky dual horsepower/torque limiting governor would've come up at some point during the testing.

Thanks for the help everyone!

 

[ivymike]

I know a bunch of people at Cat, if you have a serial number for the misbehaving engine I find out what the governor type was when it left the factory.

 

[Fabrico]

Assuming you are lowering the no load speed to 1900 by backing off on the throttle, you have not stated what you do with the throttle as you apply more load. I'm not highly familiar with that particular pump, but some will not reach full fuel on part throttle. I believe catserveng is spot-on in pointing out the need to obtain rack measurements.

 

[gruntguru]

Fabricio.
High idle is full throttle, no load i.e. max speed of the governor. The procedure from there is to apply dyno load to reduce the rpm so the numbers are all WOT.

Braddles.
Sounds strange. Have you tried this with a different engine?, Could there be hysteresis in the torque measurement system? Extra braking torque is required to overcome engine inertia when loading down from higher revs and perhaps the dyno is holding the higher reading for some reason.

Engineering is the art of creating things you need, from things you can get.

 

[Fabrico]

Gruntguru
High idle is full throttle, no load i.e. max speed of the governor. The procedure from there is to apply dyno load to reduce the rpm so the numbers are all WOT."

gruntguru You might try digesting what the OP said in his original post. "We then drop the engine to 1900rpm, no load, increase the load, and we find we can only reach about, say, 325 Nm until the engine speed starts to drop off."

This points to precisely what I outlined.

 

[gruntguru]

Sorry. Mis-understood your question. I think it is safe to assume they go back to full throttle when re-applyiong the load, otherwise his question is ridiculous.

Engineering is the art of creating things you need, from things you can get.

 

[redlinej]

The guys that was testing the engine was correct ,most turbo charger make power as the load increase.

 

[gruntguru]

No way this much difference (650 Nm vs 75o Nm) is normal. Turbo diesels run plenty of excess air and a small reduction in boost will not cause the injection system to reduce fuel at all, result being no performance drop either.

Engineering is the art of creating things you need, from things you can get.