|
|
spdracerut posted on October 20, 2011 00:35 
Turbo Tech: Turbo Troubleshooting
By Khiem Dinh
FACT: most of the time, the turbo is not the source of the problem. One caveat to this statement, this only applies to turbos made by one of the major OEM turbo manufacturers (they are really the only ones capable of properly engineering turbos): Honeywell Turbo/Garrett, Borg Warner, MHI, IHI, and Cummins. Out of all of the turbos that get swapped by dealers and shops (think diesel big rigs, work trucks, construction equipment, etc), an average of 90% of them have absolutely nothing wrong with them. Of the remaining 10%, about 5% or so of the failures were due to user error. Basically, the dealers see a performance issue and jump to the conclusion of a faulty turbo, and they would be wrong. Also, let me say this up front, turbos don’t have seals! I’ll explain this more in-depth in a bit.
Remember that turbos are pretty simple devices with the main components on a basic turbo being: compressor wheel, turbine wheel, a shaft that connects the two, and the bearing system. They get a little more complicated with the addition of variable nozzle turbine parts, but still pretty basic stuff overall. Exhaust goes in to turn the turbine wheel and flows out. The turbine spins the compressor which sucks air in and pushes it out. Oil goes in to support and cool the bearings and flows out. Water goes in for cooling and flows out. As they say, the devil is in the details which separates the well-engineered turbos from the bad turbos (and probably not engineered at all), but if a turbo checks out fine on the end-of-line quality check, it's probably fine.
 |
| Here's part of a guide you can find at TurbobyGarrett to help troubleshoot issues on turbocharged vehicles. |
The Aftermarket world is a bit different from the OEM world though, in that you have a bunch of hacks that supposedly know what they're doing working on customer cars. There are many many (many) sources of performance problems as you can see in the Garrett Troubleshooting Guide, but let me go over a couple real world examples I've dealt with.
A so-called tuner somehow got a hold of my work number and was completely flipping out. He had swapped out the stock turbo on a vehicle for a bigger one and wasn't seeing any performance gains. So of course, he assumed something was wrong with the turbo. The situation was that he was running more boost than he was on the stock turbo, yet only making the same power and even less. So I asked about some basic things like checking for boost leaks, etc. He said everything checked out. After not really getting anywhere with him, I asked for datalogs. So I go through the datalogs and compare MAF readings. The bigger turbo was flowing 30%-40% more than the stock turbo which was clearly in the datalogs. I dug a little deeper and checked the timing and found that the values were extremely low. This was super simple data analysis and troubleshooting that the tuner should have done instead of freaking out. Higher MAF readings equaled proof the bigger turbo was indeed flowing more air. Low ignition timing explained the low power. So I made some pretty little charts in Excel to show the issues and emailed the info to him. I never did hear back from the guy. Funny thing though, a few days later on a forum, a customer talks about how good of a job the tuner did on their car...
Probably the most common issue we hear about in the Aftermarket world is oil smoke. This is where we get into the issues of 'seals'. Turbos don't have a seal in the way you might commonly think of seals; it's not like Tupperware where you close the lid and you get an air-tight seal. Turbos use a combination of piston rings and other tricks to try to keep oil in the center housing and out of the compressor and turbine housings. Piston rings in turbos are a type of dynamic seal which sits between the rotating shaft and the stationary center housing. They help prevent oil from leaking past but they are not a 100% seal, so oil does get by them. The cause of oil leaking past the piston rings is pressure.
 |
|
The piston rings, or 'seals' don’t really go bad. There is the very slight chance they can be damaged during installation, but that's very uncommon. If you think the 'seals' are bad, then you most likely have a much bigger problem which is bearing system failure. If the bearing system is on the way out, it allows the shaft to wobble a lot, leading to piston ring damage and wheel rubs. Turbos employ a couple other techniques to create oil slingers. The goal of the oil slinger is to fling/sling the oil away from the ends of the shaft. In the case of the turbo in the picture, two grooves are machined that act as the slingers. The reduced volume of oil at the ends means less oil can leak out.
|
There are basically three areas of pressure in a turbo: compressor side, turbine side, and within the center housing. If the pressures in the compressor and turbine housings are greater than the center housing, then oil will not leak out. If the pressure in the center housing is greater than the compressor and turbine housings, then oil may leak past the piston rings.
Thursday, October 20, 2011 7:56 AM
What's your opinion or advice on rebuilding turbos with bad bearings? I wore out an HKS turbo and I was thinking of rebuilding it for a backup. Two different shops I talked to said they didn't advise it, but I'd like to hear from a real engineer on the subject.
Thursday, October 20, 2011 10:30 AM
HKS used Garrett ball bearing CHRAs (center housing rotating assemblies) on their older turbos, so I'm assuming that's what you have. They are not rebuildable. I could personally take one apart and put it back together, but I have no way to balance it properly and that is the key to reliability. All the major turbo manufacturers use very high speed balancers on these smaller sized turbos, think 100k+ rpms. The really really small turbos are spun up to 200k rpms. Also, there are very specific specifications on how to assemble the compressor wheel onto the shaft and torqueing the nut. The method on installing the wheel and torqueing the nut actually have a significant affect on the balance, so doing it correctly is important to reliability. So anyone trying to 'rebuild' a CHRA is basically compromising reliability.
Thursday, October 20, 2011 10:36 AM
I should say, rebuilding a ball bearing CHRA. Journal bearing turbos are easier to rebuild. Also, the larger turbos, i.e. GT45+, low-speed balancers are fine. Like in diesel truck pulling/sled pulling, they'll shread through a turbo each run. They literally bring like 6 spare sets of turbos per event. A lot of those guys like to use journal bearing turbos because they are cheap and can be rebuilt at the local corner diesel truck store. These are the really big sized ones, GT45-55 etc, where the low-speed balancer is acceptable.
Thursday, October 20, 2011 10:54 AM
Great article Khiem. There are lots of problems that can happen on the control side as well and every variable needs to be looked at in order to really get to the root cause of any problem. On the note of off-market turbo makes, we had a customer that was putting the blame to one of our external WGs as the cause for not making boost. We brought the car in and put it on the dyno to have it in a controlled environment. We put 100 psi to the top of the WG with a lift sensor to keep it closed, Run 1 = 10 psi, Run 2 = 5 psi, Run 3 = 1 psi. This is with the wastegate closed no boost leaks on a supposedly "freshly" rebuilt turbo. It still amazes me how little attention people put to watching exhaust manifold pressures. Turbos are pretty simple in fact, pressure in, pressure out and shaft speed. Great comments also on how quality in turbos is all about manufacturing capability and tolerancing. Tolerancing = $$$$$.
Thursday, October 20, 2011 1:47 PM
Awesome article Khiem. I have a Tech question for you. My SR runs great. Runs hard- 10PSI with T04Ecom./T-31 turbine. Solid 10 PSI all the way to redline. Tial external wastegate. Useing a Profec-B, I cannot get the car to boost any more than 10 PSI.(spring pressure in WG). I tested the controller; solenoids will cycle with pressure applied to reference port on controller. can hear solenoids cycle off/on. Checked and triple checked my hose routing to WG. It is correct. Switch settings on back-correct. What would the next test/check be? maybe just disconnect the hoses to the wastegate completely(to see if it will go over 10LBS)? But that sounds a little sketchy. What would be the proper way to continue diagnoising this issue? Thanks Khiem! Or should I start a thread in the Forums with you??
Thursday, October 20, 2011 1:47 PM
Obviously, there are people in every repair industry that don't know what the fuck they're doing, but I resent you branding heavy duty diesel techs as parts swappers. Who are you and where are you getting these statistics? What experience do you have in this field? I have yet to see someone I work with replace a non-VNT turbo in an attempt to fix a low power complaint, and don't get me started on the reliability of VNT turbos in the field, or the amount of low power RO's I get for engines equipped with them.
Thursday, October 20, 2011 2:24 PM
Don't take offense as I'm not 'branding' the diesel industry, I used diesel as the example because they make up the majority of the turbos out there. I'd guesstimate the current turbocharger market at around 20 million units a year. Where do you think most of those are going? I can tell you it's not gasoline passenger cars. Where do I get my stats from? Recent training I had from one of our quality assurance and troubleshooting guys that has been tearing down turbos and handling warranty/field issues for over 20 years. The two survey periods were over the ten year periods covers the 1990's and 2000s. Over those two ten year periods, the stat remind roughly the same, 85%-90% of the turbos returned to us for warrenty check had absolutley nothing wrong with them. It sounds like your shop is doing great work and is legit. Let me ask you this, how many dealers or shops would you take your everyday car to for checking out an issue? I can tell you there's not a single shop that I would take my car to. The ONE time I've taken my car to the dealer was because I needed a record of damage to my car from poor road coditions (alignment got knocked waaaaay out), so I had the car dealer check the alignment and give me a printed record. Of course, in the course of checking the frame and suspension for damage, the monkey working on the car decided to use an impact wrench on the lugnuts when putting the wheels back on which way overtorqued them; considering how hard it was to take them off, I'd guess around 150 lb-ft when the spec is 80-85. And have a friend where the dealer forgot to put the oil cap back on the valve cover after an oil change and the whole engine bay got sprayed.
Thursday, October 20, 2011 2:38 PM
Actually, now that I think about it a bit more, the turbo market is probably closer to 25-30 million units a year. I'd guess maybe 1-1.5 million of those are gasoline? So that leaves the rest on diesel applications: trucks, cars, buses, stationary power generation, 15L engines used on earth moving equipment, construction equipment, ocean liners, etc.
Friday, October 21, 2011 12:56 AM
What is the frequency of a cracked turbine housing actually causing an issue? Off the top of my head, I have seen IHI turbos from Subarus with cracked housings. Most of the pictures, and the one I saw physically, had small cracks either by the waste gate outlet or turbine outlet. It "appeared" common, and also appeared to be one of those things that forums blow out of proportion, as I did not hear first hand of anyone that had a failure related to it.
Friday, October 21, 2011 2:24 AM
I would like you to specify, do those stats include VNT turbos?
Friday, October 21, 2011 11:11 AM
@sr20freak, I ran into a similar issue a long time ago with a Greddy Profec. IIRC, on the solenoid, there are three ports; one from the signal, one going to the wastegate, and the third port which needs to vent to atmosphere. My friends and I thought that third port was suppose to be plugged, but it needs to be open in order to bleed off the air 'signal'. If it can't bleed off, then it goes to the wastegate just the same as if the solenoid wasn't there. So try that. If that doesn't work, then I don't know.... @binary, small cracks are really not an issue as long as they don't create a leak path to the outside of the turbo. On gasoline turbos, it's practically impossible to prevent all cracks due to the temperatures. But for sure, no cracks on the exterior are allowed. The gasoline turbos, especially from the 90's on DSMs and SR20 would crack around the wastegate port because of the big changes in wall thicknesses, sharp edges, and cheaper materials back then. Gasoline turbo cars also use to run cooler back then (~900C). Modern direct injection gasoline turbo cars are running well over 1000C now, so the housing materials have to be much better. With divided turbine housings, cracks in the divider wall CAN be a much bigger issue if it causes chunks to fall out, but often times the cracks just go straight outward and don't intersect each other. It's also nearly impossible to keep a divided/twin-scroll housing from not cracking in the divider wall. @DieselTech, those stats include the entire portfolio of turbos. What I'm not clear on, but it would make sense, is that those stats do not include turbos at the end of their service life. Can't blame a turbo for failing if it's already gone it's service life like you can't blame a wheel bearing for failing after 150k miles. But I don't know that for sure. I'd say VNTs are around their 3rd generation of development or so. It was before my time, but I seem to recall stories of issues on the first gen stuff on applications like Powerstroke, DMax, Cummins, basically the US light duty truck market. Keep in mind that every diesel car in Europe (and around the world) has a VNT and diesel sales have outnumbered gasoline sales in Europe for the last half decade. In 2006, about 16 million cars were made in Europe, so figure half of those were diesel. So 8 million diesel VNT turbos on cars, not including all the work trucks I don't think. I'm going to start pulling numbers out of my arse here, but let's say out of the 8 million of those cars, 10% go in and get turbos replaced, so 800,000. Of those, 5% are actually messed up, that's 40,000 messed up turbos. That's still a lot of messed up turbos, though percentage wise when looking at the entire sales volumes of turbos, it's relatively low. The fact remains that many places do improper troubleshooting analysis and assume it's the turbo that's the problem. Heck, one of the guys here in the lab that has been working on turbos for 20 years or so replaced a turbo on our DMAX lab truck thinking that was the issue. Turns out it wasn't, and it was the EGR system or something. VNTs are trickier to diagnoise as you can get vane sticking issues, and issues related to wear that would have to be found. You can't juse check shaft play and call it good/bad as on a regular turbo. Regardless, the lesson hear is that all the subsystems from the air filter to the exhaust tip need to be checked. One more story about improper analysis (not related to a turbo), but a friend took his car to the dealer two weeks ago for a clutch issue with the pedal going to the floor. The dealer thought, oh, must be the hydraulic clutch line, so they replace it and give the car back. Pedal still goes to the floor. So now the car is back at the dealer and they're swapping out the master cylinder. It's a Nissan dealer BTW, and that's trying to trouble shoot a system much much simplier than the powertrain system.
Friday, October 21, 2011 1:24 PM
I understand that the point here was to make it clear that proper troubleshooting is required, but the fact remains that VNT turbos DO cause performance issues, both at the end of their service life, and in some cases, well before the end of their service life, and I feel like that was not made clear. I will say that, after thinking about it more, alot of the in-warranty VNT issues I deal with would be with reman'd turbos, which I would assume the turbo manufacturer does not get involved in, but we do still see new turbo issues. And as far as dealer (and indies too for that matter) horror stories go, yes, everyone's got one, but seriously, we are not all like that. I take alot of pride in my work, and work/have worked with alot of people that are the same way. I will share one good story though. A customer brought a truck in with 400,000 miles (Detroit Series 60 that uses chassis air pressure for VNT operation) that had a shattered compressor wheel, and turbo overspeed codes. The customer said he almost knew it, as it would be his 5th turbo since the truck was new. I asked him where he was getting these turbos put on at, and he said the local Freightliner dealer. Apparently they did not feel it was necessary to troubleshoot the overspeed codes. This unit was equipped with an air system that required the primary system to be full before the secondary system (where turbo gets air) would fill, as well as a tag axle that required nearly full secondary air pressure to lift completely off the ground and basically doubled the size of the secondary air system. This, combined with numerous air leaks that left his tanks empty each morning, and a driver that put his foot to the floor every morning to build air faster, resulted in extended overspeeding at the start of every day. The driver didn't worry about the CEL's though, as they always went away after he built air pressure.
Friday, October 21, 2011 6:35 PM
DieselTech, there are the handful of good shops/places/etc and they are the ones that will get the repeat customers. There is only one tire shop in a 10 mile radius that I trust my car to and there's a LOT of shops in a 10 miles radius in LA. When reinstalling the wheels, they actually thread the lugnuts on by hand, snug them up with a air gun set to a very low torque, and then use a torque wrench. Road Race Engineering (RRE) is another group of guys I would trust; Mike Welch is a badass and works his butt off. Before I go off on a tangent, reman'd turbos are bad news. I've taken apart a brand new DDC S60 reman and basically all they did was remachine the shaft and center housing to accept new bearings. The bearing system is very carefully designed to handle shaft motion, etc. So by them modifying the design and using unknown clearances, reliabilty is thrown out the window. It looked like they reused the thrust bearing. Worse yet, they reused the entire VNT system without replacing a thing. It didn't even look clean as it had slug and crap on it. The main crank arm that rotates the unison ring showed significant wear as did the unison ring itself putting both out of tolerance. The compressor wheel was also damanged with a bent blade tip. The VNT parts of this DDC S60 reman had the equivalent wear of 140k miles of wear. So this brand new reman turbo starts life like it has already been used 140k miles. It just BARELY passed our end of line check. Which means it was on the verge of going out of spec. The reman also showed significantly lower turbine efficiency, like 8% points which really kills fuel economy. My advice would be to never use a reman, they'll cost the customer more in the end. On to the tangent, I was meeting up with a buddy of mine whos an aspiring actor (it is LA). He recently got a job as a host at a restuarant and has already been promoted twice and is now training others, and he's only been there TWO MONTHS! Why? Because no else there gives a shit about their job. I worked in a grocery store when I was in high school and I was the best bagger there was because I actually cared. When I sold shoes at Just For Feet in college, I had repeat customers because I cared. It annoys the hell out of me when I try to make it as easy as possible on the bagger at the grocery store (I pre-arrange cold stuff together, boxes together, cans/bottles together on the conveyer belt) and they still can't bag it correctly. Seriously? Super simple job and they can't get it right. A lot of people just don't care which is why maybe a bunch of our jobs have gone overseas. I guess on the flipside, it gives us people who do care better job security. I read a story on CNN about an immigrant up in Minn. looking for a job he could do. He saw all the storefront windows so he started working as a window washer. The guy will put in 16 hour days during the busy season and makes like $150k a year busting his ass washing windows. Not glamorous, but he does care about his work and is getting rewarded for it. I guess the point is, no matter what the job, those who put in the effort, do their best, and take pride in their work will succeed. End of tangent :)
Saturday, October 22, 2011 1:36 AM
Holy. Shit. We like to joke about Detroit's remans, but I did not realize it was actually that bad. Isn't the rated service life for one of those only 300k to begin with? That would definitely explain two recent trucks that both came in VERY shortly after getting reman'd turbos installed, with the same low power complaints they had to begin with. I watched a couple of good techs pull their hair out trying to find issues that weren't there until they finally got the ok to try another turbo. Their injectors and cylinder heads also give us alot of issues. Was it just the one turbo you tore down? How long ago was that and why? Detroit's not the only one with reman issues. I remember Navistar's long blocks. Absolute jokes. Cummins never seemed to give us many issues, though I never was as heavily involved with them. I will say that Allison's reman'd (retrans as they call them) transmissions are top notch, and come in badass shipping containers.
Saturday, October 22, 2011 8:33 PM
If Allison is reman'ing their own trannies, then I'd expect them to be fine. Like my dad buys refurbished Sony stuff and has never had a problem with them. Send me a PM with your email. I have some more info.
Sunday, October 23, 2011 12:48 AM
Khiem, You stated something that caught my eye when you said to walk away from turbo kits that drain to the oil pan. I am running a Greddy Turbo kit with all of the "fixes" and a upgraded Greddy T618Z Turbo by AMR that adds a water and oil cooled CHRA (comes from Greddy only oil cooled). However with this kit, draining back to the pan is the way Greddy designed it. I am curious what your thoughts are on this and if you would have a better suggestion. Thanks, Scott
Sunday, October 23, 2011 8:22 PM
Scott, the majority of kits that convert a non-turbo car into a turbo car probably drain into the bottom of the oil pan, through the drain plug location, to make the kit as simple and bolt-on as possible. For many people, that's probably okay. From a marketing standpoint, if kit A is a bolt-on affair and kit B requires you to drop the oil pans and drill out the block, a whole lot of people will go for kit A due to ease of install. The best analogy I can think of is a slow draining sink. If the faucet is only turned on 50% and the drain can keep up, then it's fine. If the faucet is only cranked open 100% for short durations, it's fine. But when that faucet is open 100% for extended periods of time, that sink will overflow. It's very much like the brakes and stock oil cooler system on the S2k (and pretty much any car sold). They're perfectly adequate for short durations of hard use. But they are inadequate for sustained hard use. Like the 370Z is perfectly fine to to drive around in and do the occasional WOT romp on the highway on-ramp. As soon as that car sees 3 laps on a road course in warm weather, it overheats the oil. What percentage of owners will drive the 370Z hard enough to cause a problem? Not many which is why the car doesn't come with a bigger oil cooler stock. But for those that do drive hard, the stock system is inadequate. Anyways, the proper drain location for any turbo setup on any vehicle is above the oil level in the pan. That way, oil can not back up in the drain line.
Tuesday, October 25, 2011 7:58 PM
Undersized oil drain tubes are truly a plague. I constantly hear people on various forums badmouthing quality turbo companies for "being unreliable junk" or "having crappy seals" because idiots are installing them with tiny -10AN drains and feeding them 80+ psi of oil pressure with -4AN feed lines. Some are so thickheaded that they realize they have an oiling problem and fix it by excessively restricting the oil supply rather than enlarging the drain as they should. The turbo manufacturers actually have spec sheets that state the oiling requirements for each turbo... not a bad thing to read before planning a turbo swap. This past summer I swapped out the stock twin turbo setup from my Supra and installed a Holset HY35W. A -3AN feed line with a -12AN drain tube that drains to the stock location on the block should keep it smoke free for a long time.
Monday, November 14, 2011 4:34 PM
I
Monday, November 14, 2011 5:23 PM
Thanks, great article and I really appreciate the feedback.
 
|
|
MotoIQ Proudly Presents Our Partners:
|