CR engine HPFP analysis

[CedarPark68]

Make sure to read the latest letter form the NHTSA to VW. ( October 7, 2011 )


http://www-odi.nhtsa.dot.gov/acms/docservlet/Artemis/Public/Pursuits/2011/EA/INPR-EA11003-48542.pdf

 

[sgoldste01]

Make sure to read the latest letter form the NHTSA to VW. ( October 7, 2011 )


http://www-odi.nhtsa.dot.gov/acms/docservlet/Artemis/Public/Pursuits/2011/EA/INPR-EA11003-48542.pdf

Very interesting document. My favorite part is #31 on pages 15-16.

 

[woofie2]

Very interesting document. My favorite part is #31 on pages 15-16.

I wonder if those oscillations were feed back from the #2 injector line vibrations?

 

[oxford_guy]

Is corrosion a factor in failing HPFPs?

I don't know. I was just following up on your post about titanium. However, the 2005 article said a coating problem caused failure for that design. I assume that means the coating was worn away. I don't know if the coating that was used is more robust than an iridium surface of some sort (alloyed, or something).

Iridium is supposed to be very hard and very corrosion-resistant. The problems with it, that I know of, are its brittleness (osmium may be more brittle) and high cost. Rhodium, which is super-expensive, is also used in alloy high-temp corrosion-resistant coatings, according to Wikipedia. It's used in catalytic converters.

 

[kjclow]

The rhodium is used in the catalytic converter as a catalyst to react with the exhust gasses. There is no mechanical wear required. I don't know how well it would hold up as a material for the HPFP. Lots of hard corrosion resistant metals work well in surface coatings but these will not have any mechanical wear. Most of these would also be cost prohibitive to ever use for this type of application.

 

[Honeydew]

You guys are discussing some very rare & expensive and/or not very wear resistant alternative materials here. Tribology is a rather complex branch of materials science and apparent solutions aren't necessarily the ones that will work or the most practical.

 

[dweisel]

You guys are discussing some very rare & expensive and/or not very wear resistant alternative materials here. Tribology is a rather complex branch of materials science and apparent solutions aren't necessarily the ones that will work or the most practical.

The nicasil coating is the recomendation of a skilled machinist that works on Nascar engines. He stated that the pump would last forever with the bore coated with nicasil. The cost at the time I enquire was around $500.00

 

[Honeydew]

The nicasil coating is the recomendation of a skilled machinist that works on Nascar engines. He stated that the pump would last forever with the bore coated with nicasil. The cost at the time I enquire was around $500.00

My comment was referencing the recent discussion on alternate base materials rather than the earlier discussion of nicasil coating and was not intended to chide any members but prompt a more holistic analysis of the problem. I agree that nicasil would be a promising coating solution for the Al bore.

 

[specsalot]

CedarPark68 - Thanks for linking the latest NHTSA informational request. They are drilling very deeply into questions of development for this pump including references to design changes. They are also drilling into operating conditions, failure modes, and customer treatment issues. There will be no fig leaves to hide behind if VW makes an attempt to fully respond. A perfect recipe for folks under the gun in the organization(s) to break ranks.

It is good to see NHTSA called out VW for defending a free to rotate follower design while at the same time working with Delphi to develop a follower that was constrained. It shows that NHTSA folks are on the ball not just bureaucrats on a mission.

I've spent many days studying the 'failed' HPFP in my possession. With ~ 22k miles, this pump shows an absence of significant wear or defects. The use of special anti-friction / anti-wear coatings can be seen in several area's of the pump including:

• Cam shaft bearing inserts (radial and thrust faces)
• Cam follower roller carrier
• Cam follower piston
• HP pump piston

Despite the fact that there were a lot of nearly invisible metal fines in circulation in this pump, these coated surfaces seem to be intact with little or no evidence of wear. How long coating can hold up in this microswarf environment is the 64cent question. It's looking more and more like Bosch has more at stake than VW given the utilization of the CP 4.X pumps in the wider global vehicle fleet. I think the days of laying blame solely on fuel quality are rapidly coming to an end. This latest information request will certainly put a lot more "blood in the water".

 

[Bugdoc]

specsalot, interesting post. Questions: you use quotations around 'failed'. Why so? It failed or it didn't, no? And, if wear didn't cause the 'failure', then what do you think did?

 

[emdeeaitch]

I've spent many days studying the 'failed' HPFP in my possession...

Fantastic. For those of us who missed it, can you remind us which revision of the pump you have in your possession?

 

[kjclow]

specsalot, without destroying the failed pump you have is there any way to allow the roller and pistons to sit in gasoline for a few days? I know that would be the worst case scenario, but I am still wondering if gasoline effects the adhesion of the surface treatment to the roller or cams. This would kind of support VWs stance of bad fuel but once we know what aids in the initiation of the failure, perhaps it would be easier to then try to determine at what level of gasoline does the failure start and how to truely prevent such a level of contamination from occuring.

just so I don't get flamed for sounding like I support the contamintation theory: I do believe that these engines should be able to with stand a certain level of cross contamination and if a misfueling does occur, there should be an easy way to "save" the engine.

 

[eddif]

specsalot, without destroying the failed pump you have is there any way to allow the roller and pistons to sit in gasoline for a few days? I know that would be the worst case scenario, but I am still wondering if gasoline effects the adhesion of the surface treatment to the roller or cams. This would kind of support VWs stance of bad fuel but once we know what aids in the initiation of the failure, perhaps it would be easier to then try to determine at what level of gasoline does the failure start and how to truely prevent such a level of contamination from occuring.

just so I don't get flamed for sounding like I support the contamintation theory: I do believe that these engines should be able to with stand a certain level of cross contamination and if a misfueling does occur, there should be an easy way to "save" the engine.

The very small diameter roller depends on fuel viscosity a lot. The contact patch is narrow with a small roller and the viscosity needed for this contact patch is great. Anything that happens to this design is going to push toward disaster (your adhesion comment as an example) .

I tend to think (and my thoughts may not be correct) that a larger roller diameter would help some. The problem rapidly becomes the fact that: the slider diameter increase will cause the need for a shorter roller. Surely VW knows and tested all this.

The square slider bore will allow roller diameter to grow and maintain the same length roller.

Looking at the history of pumps a lot of different roller configurations have been tried in the HPFP industry. Perhaps it is about time they find the best compromise between cost and longevity. If anyone already knows which design works best, I am all ears (axle, captive roller, etc ?).

eddif

 

[tdiatlast]

A BIG thank-you to all of you that have taken on the HPFP issue with such enthusiasm. Your detailed discussion is fascinating, and I think we've all learned a great deal about the extremely refined engineering that goes into the development of such a sophisticated design.

Question: Is the detailed October 7, 2011 letter any indication as to the seriousness of NHTSA's position? I've never had any experience with NHTSA, just wondering if this is typical.

With 31k and 48k on my TDIs, I'm just attempting to predict the future here.

 

[Honeydew]

bore coating

Nicasil (nikasil) is essentially a electrolytic watts nickel strike with silicon carbide particles co-deposited. The silicon carbide particles provide wear resistance and are held together by the nickel. We have a guy here at work who is a race engine builder and a former major piston ring OEM executive. He used to have issues with rings wearing the nickel out of the nikasil matrix on coated cylinder bores. The only rings that worked reliably were plain soft cast iron, but that configuration did work well. Some fuels also have a negative effect on Nikasil.

DLC is a type of PVD coating and is rumored to be one of the coatings used in the CP4.1. These coatings are extremely difficult to apply to bores especially in a cost-effective production environment. This may explain why the piston bore is not coated like the piston is and also may explain why the touted DLC injector nozzles offered by vendors here have a DLC coated plunger but the nozzle bore is only "hardened".

Another coating option for the Al bore is silicadizing. This is a rather uncommon coating but is well suited to this application. There are varieties applied using Direct Current and Alternating Current and the AC variety would be the preferred type for this application.

 

[Niner]

An interesting read on our injectors and fuel pump, and the effects of fuel quality on wear...


http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA547468&Location=U2&doc=GetTRDoc.pdf

Brief summary, 200 hour run time, wear and scuffing already on the pistons, one roller showing wear into the saddle finish at the side, unexplained by the folks that observed it, but deemed "acceptable". Me thinks a 1000 hour test is required. Note that the lubricity of the 4 fuels is all over the charts, above the limit. Makes you wonder about Bosch's claim about fuel lubricity being valid or not?

 

[Niner]

Nicasil (nikasil) is essentially a electrolytic watts nickel strike with silicon carbide particles co-deposited. The silicon carbide particles provide wear resistance and are held together by the nickel. We have a guy here at work who is a race engine builder and a former major piston ring OEM executive. He used to have issues with rings wearing the nickel out of the nikasil matrix on coated cylinder bores. The only rings that worked reliably were plain soft cast iron, but that configuration did work well. Some fuels also have a negative effect on Nikasil.

DLC is a type of PVD coating and is rumored to be one of the coatings used in the CP4.1. These coatings are extremely difficult to apply to bores especially in a cost-effective production environment. This may explain why the piston bore is not coated like the piston is and also may explain why the touted DLC injector nozzles offered by vendors here have a DLC coated plunger but the nozzle bore is only "hardened".

Another coating option for the Al bore is silicadizing. This is a rather uncommon coating but is well suited to this application. There are varieties applied using Direct Current and Alternating Current and the AC variety would be the preferred type for this application.

Someone look into silicadizing an aluminum bore, and if that's what Chevy used on the 1970's Vega motor, that had an aluminum block. It didn't last very long, at all.

 

[manual_tranny]

That is a very appropriate warning, Niner. If I may add: while I DO lack much of the knowledge necessary to make an appropriate judgement on the matter, I would expect that metallurgy and coatings may have made some progress in the last 40 years. I would hope that a Nikasil/silicadized coating might work better today than those used in the '70s Vega. Pardon my speculation, but perhaps there is a better application technique and/or alloy mixture in 2011 than 197X?

 

[truman]

Porsche used the siliconized block in one of their models- maybe the 928 IIRC.

 

[specsalot]

specsalot, interesting post. Questions: you use quotations around 'failed'. Why so? It failed or it didn't, no? And, if wear didn't cause the 'failure', then what do you think did?

Without running it on a car I can't prove its good. Nothing I've seen suggests it won't work. The original owner has not provided the VW DTC's that were the basis of VW's decision that his entire fuel system needed replacement as his expense.

Based on the narrative of the original owner it appears there his tank had water contamination and lots of rust. He probably was in the process of losing his Aux Fuel Pump not his HPFP. No way of knowing, only speculation at this point.

 

[Niner]

That is a very appropriate warning, Niner. If I may add: while I DO lack much of the knowledge necessary to make an appropriate judgement on the matter, I would expect that metallurgy and coatings may have made some progress in the last 40 years. I would hope that a Nikasil/silicadized coating might work better today than those used in the '70s Vega. Pardon my speculation, but perhaps there is a better application technique and/or alloy mixture in 2011 than 197X?

Agreed, on progress... No problems, it's all good, in the sense that it might very well be the solution to the problem with these pumps.

 

[specsalot]

Fantastic. For those of us who missed it, can you remind us which revision of the pump you have in your possession?

I don't have the build date for the vehicle, but believe it's an early 09 build.

 

[specsalot]

The very small diameter roller depends on fuel viscosity a lot. The contact patch is narrow with a small roller and the viscosity needed for this contact patch is great. Anything that happens to this design is going to push toward disaster (your adhesion comment as an example) .

I tend to think (and my thoughts may not be correct) that a larger roller diameter would help some. The problem rapidly becomes the fact that: the slider diameter increase will cause the need for a shorter roller. Surely VW knows and tested all this.

The square slider bore will allow roller diameter to grow and maintain the same length roller.

Looking at the history of pumps a lot of different roller configurations have been tried in the HPFP industry. Perhaps it is about time they find the best compromise between cost and longevity. If anyone already knows which design works best, I am all ears (axle, captive roller, etc ?).

eddif

Interesting comments.

I'm shopping for some better measuring tools to try to capture the actual clearances around this roller. I have crudely estimated vertical movement at 0.002" to 0.003" which yields diametrical clearance of ~ 0.001 to 0.0015". Up and down movement and the rotation should provide for exchange of lubricating fuel around this roller. The pump seems to operate with a wide range of fuel supply temperatures, despite the pre-heating valve in the fuel filter. Through in some significant changes to roller loading based on CR pressure changes and you have a recipe for lubrication problems. And on top of this lets not forget the micro swarf in circulation. At ~ 22k miles, I would have expected to see some kind of wear on the coating of the roller carrier. But it looks pristine.

Can't say what's best for automotive, but on large marine engines all the fuel cam followers I've ever seen utilize a roller that turns on an internal axle. Likewise, every follower design I've seen has been constrained to prevent rotation. First glance at this HPFP design set off alarm bells in my head.

 

[manual_tranny]

From another HPFP related thread and from Delphi: A rather damning bit of evidence against the design of the Bosch pump:

"The roller cam shoe mechanism has been optimized for reduced torque, dynamic mass, and noise via the Delphi patented static shoe guide. The component is pressed straight into the housing so that the shoe guide, different from previous solutions, does not oscillate during pump operation keeping the dynamic mass low, permitting a small plunger return spring size and optimum packaging geometry. The design safely prevents the shoe on the DFP6 pump from lateral rotation which eliminates the risk of fatal pump damage and allows for a mass of just 2.4kg."

Here is the source: http://delphi.com/about/news/media/pressReleases/pr_2010_05_11_001/

I do not claim to know the physics involved inside the Bosch or the Delphi pump, but I have been meaning to share this information in this thread for a couple of days.

 

[specsalot]

Not detailed in my pic above, there was some coating wear on the ends of the roller where it would be contacting the follower body. One side more than the other. Forgot to mention that. In the area where failures occur, no coating wear can be seen.

 

[eddif]

Not detailed in my pic above, there was some coating wear on the ends of the roller where it would be contacting the follower body. One side more than the other. Forgot to mention that. In the area where failures occur, no coating wear can be seen.

This all gets interesting

While the curved, half circle roller support lost part of it's coating, the end of the roller (under the 9:33 PM) did not seen to lose it's coating.

The inside of the slider (on this pump) does not show a wear area like the military tested slider.

eddif

 

[oilhammer]

I just have to say, I am proud to be amongst such helpful folks and forward thinkers. Shame Bosch does not have these same minds amongst their ranks. Keep up the good work, guys, and what an excellent discussion.

 

[torqueit]

An interesting read on our injectors and fuel pump, and the effects of fuel quality on wear...


http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA547468&Location=U2&doc=GetTRDoc.pdf

Brief summary, 200 hour run time, wear and scuffing already on the pistons, one roller showing wear into the saddle finish at the side, unexplained by the folks that observed it, but deemed "acceptable". Me thinks a 1000 hour test is required. Note that the lubricity of the 4 fuels is all over the charts, above the limit. Makes you wonder about Bosch's claim about fuel lubricity being valid or not?

The lower lubricating properties of JP8 are a very well known characteristic of that fuel and one that diesel engines modified to meet military applications must contend with.

In many cases, fuel pumps had to be modified with improved coatings to prevent failure when the first transition was made between DF2 and JP8 back in the early 1990's. The attached paper is expecially interesting because at this time, common rail engine designs are just starting to be considered for military use.

I'd be inclined to agree that 200 hours is clearly not long enough to establish a clear position on the suitability of these engines and their respective fuel pump designs but I'd be very curious to see how our Bosch pumps stand up to fuel with similar fuel lubricity problems. I'd guess very poorly. This would seem to point to evidence that the Ford Powerstroke that was tested has a substantially more robust HPFP design since it did not fail catastrophically even in the face of such noticable wear.

 

[kjclow]

But they were using the Bosch 4.2. With the noticed wear, I wonder how many more hours the engine would have needed to run before failure. maybe only another 50...

 

[CedarPark68]

I simply cannot wrap my head around this design at all.

I've looked at these three pieces via all the posted photos and just do not understand what Bosch had in mind.

I like the quote back that the NHTSA put in their latest request of VW:

( at some point in time, VW said the following )

"roller is guided by the linear contact to the camshaft lobe and is therefore self-stabliizing"

"An additional guide plate would lead to a statically overdeterminated system ..... "