B7100 - white smoke from exhaust at high RPM with occasional puffs

[JohnJohnson]

Hi everyone,

Just picked up a mid-90s B7100 HST, 600 hours on the clock. I'm having a problem with white smoke from the exhaust and I'm stumped. At first it was hard to start and had really heavy white smoke. I had the injectors and injector pump rebuilt at the diesel shop. The technician suspected "fuel dilution" when rebuilding the pump, not quite sure what that means other than possible fuel contamination in the oil? It now starts right up, but still has pretty heavy white smoke for the first 30 seconds or so. Eventually it mostly clears up and becomes just a faint wisp until I open up the throttle. When I push the RPMs the smoking gets worse and it "coughs" clouds of smoke every few seconds (almost like a cylinder periodically fails to fire?) and doesn't clear up until I bring it down to idle. Am I just expecting too much? I have an older Kubota G3200 that doesn't have a hint of smoke so I guess it doesn't seem unreasonable, but I'm totally stumped on what to do next.

What I've done:
Rebuilt injectors and injector pump
Tested for proper pressure from lift pump
Put in fresh fuel.

What I plan to do next:
Compression test (possibly bad rings or valves?)
Replace fuel filter, oil filter, oil

Anything else I should check? I really like this tractor and would really like not having puffs of white smoke for something with only 600 hours on it.

Thanks for any input!

 

[twomany]

You have a cylinder that is not firing reliably.

Poor compression? Piston , rings, head, gaskets

Bad injector in SPITE of being rebuilt. (air in the line)

Malfunction of the fuel pump

 

[challer61]

I would not operate your tractor until this is fixed. Bursts of white smoke suggest water as a potential cause. Unburned fuel/oil can also make white smoke (less so in bursts) and in that case, raw fuel is washing the oil off the cylinder walls, with that associated damage. A compression test (even if positive won’t be entirely conclusive as water leaks often don’t impact compression all that much). Spray test the nozzles on a bench set up. And I’d scope the cylinders as well. Brace yourself for the idea that the head may need to come off for a gasket. Check with the prior owner to see if he had to add water or oil regularly.

 

[JohnJohnson]

Thanks for the advice. I'll do a compression test but I'll keep in mind that it won't rule out other problems like the head gasket. Is there a way to test for water in the exhaust?

One of the three injectors was clearly an aftermarket replacement and not an original, could be a poor knockoff. Makes me wonder what prompted the previous owner to replace that injector.

I'll also scope the cylinders, what should I look for? Rust, carbon?

No info from the prior owner, I bought it from a guy who owns a small engine repair shop that picked it up from an estate sale, it had sat for a while apparently. The guy who sold it to me said it had a small smoking problem that he said would be fixed with some injector cleaner, stupid of me to take him at his word. Lesson learned.

 

[challer61]

Well - Compression test may tell you about valves or rings (possible smoke source) but there is a lot of variation and could be valves or rings.

Measuring water in the exhaust is very expensive equipment. You dealer may have it but doubt it. Water leaks can also result in exhaust in the coolant. Is the coolant temp good after an hour of load? The service manual walks through this but basically, buy a radiator tester, pump it to 13psi and then let is set and see if you can find leaks. Also scope the cylinders while doing this do check for leaks. If all is good it should hold pressure for a long time. You can also add coolant leak UV dye to the water and that helps detecting water leaks generally with a UV light.

While scoping the cylinders check for scuffing or wear on the cylinder walls (fuel washing often). Check the for water leaks (under pressure) around the head gasket and valve bridge (these crack though generally only open up under temp/load but if you used the dye, run the engine until you get smoke, shut it down and you can often see dripping. Also look at the valves for burns.

The oddball injector may or may not be a smoking gun but it would require a retiming of the pump/injector. (in reality, pump/injector timing is often not set well at the factory and everyone should do it.) Someone definitely thought that cylinder had an issue.

The service manual, page 3-36 illustrates how to do a nozzle pressure test. Basically, it's a hydraulic pump and pressure gauge. Caution as the spray from the injector can cut you very badly.

You'll want to check pump pressure as well because that could also be a cause for poor injection.

If it were me, I'd also check injection timing per the manual. Once set it is usually good for a long time but that odd injector would have required a timing check which they may not have done.

If you want the service manual it is online or I can post it. The diagnostics particularly for smoke are not good but the techniques for service are very good.

By way of background, in a previous life, I designed fuel injection and piston/ring packs for medium and slow speed engines (think rail, ships, off road trucks) for all major US and EU manufacturers, travelled all over sorting out problems and wrote a bunch of diagnostic manuals (not available) for many makes and models. The Kubota is high speed in this sense but the principals hold.

I wish Kubota had cylinder head temperature gauges for each cylinder because that tells you soooo much about what is going on. Standard on most planes.

You've got a good project ahead of you!

 

[JohnJohnson]

Thanks for the response.

Can I adjust timing of a single injector? I checked the timing of the injector pump before having it rebuilt and one of the three was definitely correct and consistent but as far as I could tell the other two (including the oddball injector) were way off and pushed fuel at seemingly random times and usually after the compression stroke. When I was working on the injectors and removing the fuel lines the delivery valves for those two injectors accidentally came loose because I couldn't get a wrench in to hold them. I tried to put it back together correctly and torque it to spec, but maybe that caused my timing issues for those two injectors?

After getting back the rebuilt injection pump I made sure to use the shim that was underneath as I understand that controls timing, but I'm not sure if timing can be set individually per injector. I haven't yet rechecked the timing after having the pump rebuilt, but I will do that. I also couldn't find torque specs for the bolts holding in the injector pump.

I will try pressurizing the radiator and see if it holds pressure.

I have a copy of the WSM for the B7100 and the WSM for the 70mm stroke engines and definitely the diagnostics for smoke are not super helpful basically the only thing recommended was having the pump rebuilt which I already tried.

I will keep at it, thanks!

 

[challer61]

It appears the 70mm engine only allows you to set overall pump timing but not injector timing - no the best but works. In this case, pump pressure and injector pop pressure from the spray test will determine timing.

If you have access to a thermal camera or an IR temp gun, put the engine in high idle or under load and check exhaust manifold temps for each cylinder. One may be off - likely low.

 

[Henro]

This is a question to prove there are not stupid questions, I hope!

Just wondering if one suspected a possible gasket leak, or cracked head, if he might remove the hoses that are between the block and radiator, and then rig up some air tight covers to block the outlets, with one having a pressure gauge on it, and a valve to insert air through.

Could one pressurize the block to some level, higher than normal, like 30 psi or 50 psi, then watch the gauge over time to see if there was a leak?

Just seems like such a set up should hold pressure indefinitely, if nothing is wrong. Granted, things open up with temperature change...so maybe doing something like this would not be worth the effort...

 

[challer61]

Well, you can just pressurize the whole system and that is ok. Of course you can go as high as you like and see what leaks!

If the head is suspected base on inspection with dye in the coolant, the typical method is to pull the head and use dye penetrant.

 

[Henro]

If the head is suspected base on inspection with dye in the coolant, the typical method is to pull the head and use dye penetrant.

Understand the use of die penetrate to find cracks...but not the part about dye in the coolant to indicate something.

How does the dye in the coolant get there? OR if the dye is in the coolant, where does one look for it outside coolant passages in the engine, since the white smoke, if due to a leak, would be from coolant entering the cylinder and not visible outside the engine.

 

[challer61]

(not me by the way...)
So, you put dye in the radiator. Run engine. Check for leaks. radiator, hoses, pump, etc. AND if you can manage it, in the cylinder

 

[Henro]

AND if you can manage it, in the cylinder

Watched the video.

Obviously, external leaks would be easy. Is it even possible to see dye in a cylinder (due to a bad head gasket, or cracked head) that has been operating somewhat normally? The OP's question seems to relate to what is happening in the cylinder...

 

[challer61]

You are correct. He said he would scope the cylinders. Just remove the white light source and use the UV light or put a uv filter on the scope source. I've done it many times

 

[RCW]

Just a simple test we’ve suggested here on OTT to get an idea on head gasket state is to take the radiator cap off when cold and start it at idle.

if there’s a bunch of bubbles in the radiator, indicator of a head/head gasket problem.

Not definitive, but gives a quick and dirty indication.

 

[challer61]

Its a great test - and you can temporarily do this at load. Sometimes you need a cap extender that blocks the tractor overflow. Which can result in scalding thus I'm not comfortable recommending it. But I've done it on engines bigger than a two story house.

 

[RCW]

Makes sense.

Just thinking if no bubbles cold, he could focus more on injectors?

Just a thought.

Several times I’ve done it on my own when cold just to check.

Guess you’d call it the Redneck Checklist…..

 

[challer61]

It can work and always worth a check. However, some many head cracks don't open up until you are at temp. On big engines, the heads are pulled every year and PT'd standard. But if a leak is suspected off maintenance, you'd scope it before pulling the and taking the ship/train out of service.

 

[JohnJohnson]

How long would it take to notice coolant loss? I've maybe only run it less than 5 hours since I picked it up, the coolant level in the radiator is right up to the neck of the cap, not how big of a leak and how long it would take to notice the level drop.

 

[eserv]

This one appears to me more like a combustion issue than making steam! The first thing I would do would be a compression test.. The injection shop meant fuel dilution, the most common fuel dilution is gasoline in the fuel. Maybe a complete draining of the fuel and replace it with known good diesel fuel is in order?

 

[RCW]

Guessing it would take a while. Engine is less than 1,000cc and 15-18hp.

Maybe there's a gallon or two plus in coolant. Gallon of water makes a lot of steam...ask any maple syrup producer.

Sounds foolish, but what does the white exhaust smell like?
Any sweet smell?
Petroleum?
(Guess you can add them to the Redneck Checklist)