Choosing a Replacement Turbocharger

[GeoffBob]

It seems there is some interest in fitting alternative turbochargers, either to deliver more power (often at the expense of reliability), or as a result of a failure of the original Garrett T2. I have some experience in matching turbochargers to engines, so I thought it would be nice to create this thread for those who wish to discuss this topic. Let me say from the outset that I am not in favour of fitting large turbochargers in an effort to extract large amounts of power from small engines. I am however in favour of more reliable and more efficient technology.

For those not familiar with the principle of operation of a turbocharger I would recommend that you download a Garrett turbocharger catalog here: http://www.turbobygarrett.com/turbobygarrett/products/catalog.html which includes a basic explanation of how a turbocharger works, and how to select a turbocharger to match your engine. Garrett are the only company I am aware of that make their turbocharger compressor and turbine data freely available over the internet, which is why I prefer their products. Also note that Garrett do not supply data for their old T-series turbochargers (T2, T3 etc ...) over the internet, which have been copied adinifinitum by other manufacturers. They only supply data for their GT-series, which are vastly superior in my opinion.

If you want an old T2 or T3 (designed in the 1940's) then they still work fine, but chances are you may be buying a unit made in a less reputable country/factory, so check the manufacturers markings before you take delivery - I mean this :!: These low quality units do not perform according to the compressor and turbine data downloadable from other sources, so matching these units to your engine is generally impossible. Buyer be warned.

Having read the first few pages of the Garrett catalog and you are still serious about fitting a new turbocharger then I would recommend you purchase the book "Forced Induction Performance Tuning" by A. Graham Bell, published by Haynes in the UK, see http://www.haynes.co.uk/webapp/wcs/...10001&storeId=10001&productId=11153&langId=-1. The R5turbo engine and the infamous "Dispositif Prerotation Variable" turbo (pioneered by Jean Pierre Boudy and used to great effect on the Maxi) is discussed briefly under the heading of "Rally Engine Development". This device employs a set of adjustable vanes positioned at the inlet to a T31 compressor in order to effectively adjust the compressor inlet area and the direction at which the airflow hits the compressor wheel. The position of the vanes was controlled via the throttle linkage. So far as I am aware the Maxi employed a T04 turbine.

Another good book, although slightly less holistic in my opinion, is Corky Bell's "Maximum Boost: Designing Testing and Installing Turbocharger Systems", published by Bentley Books in the UK, see http://www.bentleypublishers.com/product.htm?code=GTUR. This book does however include a detailed explanation of the Variable Area Turbine Nozzle (VATN) turbocharger, which is similar in principle to Boudy's design, but has the variable vanes positioned around the turbine wheel, instead of adjacent to the compressor wheel (the former being a lot more tricky when you consider the temperatures involved).

For those seeking a modern VATN turbo the VAT25 (as used on the 1992 Peugeot 405 T16) can still be purchased from Garrett (I don't think it's in the catalog), and is an ideal candidate (provided you are prepared to tackle the task of operating the vanes) for the R5T 1397cc engine.

For those who have a preference, most of the calculations in "Forced Induction Performance Tuning" are in Metric units, while those in "Maximum Boost" are in Imperial units.

Comments, questions and ideas welcome.

 

[GeoffBob]

For those interested, the attached picture ("borrowed" from Classic Car Collection, hope they don't mind) shows a Maxi exhaust header with the T31/T04 turbocharger and Boudy's "Dispositiv Prerotation Variable" compressor nozzle mechanism. There is what appears to me to be an Exhaust Gas Temperature (EGT) probe to the right of the turbine flange. A very useful measurement to have in the event that your air:fuel ratio should suddenly run high (lean) as a result of a faulty fuel pump or ruptured fuel line. The warning of the sudden rise in temperature may prevent the driver from putting a hole through a piston crown or burning valves.

And how was it possible to fit the turbo in this configuration, when on all others the turbo was behind the engine above the clutch, or offest to the intake side of the engine (as on most R5T's)? Simple, the Maxi had the exhaust pipe emerging just in front of the right-rear tyre (which I think is why the fuel filler pipe and cap had to moved to the left side of the car). This freed up space behind the engine for the water-cooled intercooler employed with the Maxi and also enabled the alternator to be put back where you would usually expect to find it on a derivative of the type 840 engine (where it is quite easy to work on).

 

[Nate]

Most of the variable vane turbo's werent suited for >1bar boost... not sure if there's *any* that can handle above that...
I'd love to go down that route if there's suitable versions!

 

[GeoffBob]

Looks as if the VATN Aerocharger by Aerodyne will do about 18psi (1.2 bar) boost, but spools up in about one third of the time of a GT25. Not much more than you are looking for I know, but I think that's about the limit of the technology at the moment. For most 0.6 to 0.8 bar boost is sufficient , but I agree that for more powerful applications (such as where high boost is required on a smaller engine) this is not enough.

 

[r5gordini]

Really interesting stuff Geoff - It's the kind of thing that I would like to see posted in the Renault 5 Wiki that I run on my main site. I'm sure you won't mind if I copy it across there!

I dream about this sort of stuff and would love to get stuck in. However, current funds and project cars available mean it's out of the question. Our T2 is too precious to start messing with, and my Gordini Turbo is "mostly" original.

Thanks,

Andrew

 

[GeoffBob]

Andrew, feel free to copy any of my posts over to your Wiki. However, please note that some of my posts are based on opinion, and others on that which I understand (to the best of my ability) to be fact. I'd recommend leaving out that which is blatant opinion.

It is interesting to note that the new(ish) Porsche 911 GT2 (the 997 variant) uses two variable vane turbochargers by BorgWarner, who bought KKK in 1998. I'd be interested to know the details of what materials they are using in the turbine, but they are not telling (I would speculate that the vanes are made from inconel at least). This link http://www.autozine.org/technical_school/engine/tech_engine_3.htm#VTG explains variable vane turbo's quite nicely.

 

[Nate]

(I would speculate that the vanes are made from inconel at least).

i thought all turbo internals were Inconnel already?
although the super cheap ones are probably mild steel recycled from romanian tanks...

 

[GeoffBob]

Sorry Nate, It wasn't my intention to imply that the turbine wheel or variable vanes could be made from anything less than Inconel (although I suspect the chinese might have tried plastic). My point is that at the very least they would have to be made from iconel. Keep in mind that there are various grades of inconel with differening quantities of Nickel, Cobalt, Chromium,Titanium and Molybdenum. I think the predominant variation is on the nickel content and the cheaper alloys have no molybdenum (but I'm no expert on this). Personally, if BorgWarner used one of the Inconels, I 'd like to know which one. Data for various high temperature metals, including grades of Inconel, can be found here: http://www.hightempmetals.com/technicaldata.php

 

[harry]

Only the turbine side would need to be made of inconel, the compressor wheel can be aluminium. The variable DPV unit has rubber vanes in a wing profile. I'll up some pics of the internals. I've got one around.
Also the turbo of a maxi is a bit different from the one on the photo. The exhaust part has a tangential divided housing. If anyone stumbles across one, I'd love to borrow it to do a 3d laser scan of it, so I can reproduce at least one.

 

[GeoffBob]

For days I've been trying to remember the name of a material one better than inconel, as used in the old F1 turbos and on the Maxi5 - and then it finally came to me - Marchem steel. Originally developed for military only applications.

Harry, in case there is any confusion, as identified in my first post of this thread, there is a big difference between the DPV turbo and a modern variable vane turbo. I know of no single vehicle that today uses anything even similar to Boudy's original DPV device, where the variable vanes are mounted on the compressor side of the turbo. Here I think it goes without saying (and is understood by all so far) that temperature is not an issue and there is no requirement for high temperature materials. Hence the rubber vanes used with the DPV.

The modern variable vane turbo, the current point of discussion, is however quite another story. In this case the variable vanes are mounted around the turbine wheel, and thus both the turbine wheel and vanes are required to be manufactured from high temperature materials such as inconel.

I'd be most interested in pictures of the DPV, thanks.

 

[GeoffBob]

It seems that the development of higher temperature turbine wheels is a hot topic at the moment, 'scuze the pun, given the drive towards smaller capacity engines producing more HP per cylinder (at the expense of higher temperature) than before. For those interested, have a look at http://www.mhi.co.jp/technology/review/pdf/e453/e453001.pdf. Mitsubishi Heavy Industries (MHI) have published a paper on the first production turbocharger for small petrol engines with a turbine made from Mar-M247 (previously reserved for racing engines only, as used on the Maxi5 and other F1 turbochargers of the time). Mar-M247 is a Nickel based heat resistant alloy good for 1050'C. Sounds like a better grade of Inconel to me? I think "Mar-M247", got corrupted to the name "Marchem" in the popular literature I was reading at the time.

The very clever Japanese people at the MHI Nagasaki Research Centre have also developed a light-weight Titanium-Aluminium alloy turbine wheel with large amounts of Niobium and "other trace constituents" (that's code for secret ingredients), that will outperform both Inconel 713C (most widely used in turbochargers) and Mar-M247. The secret to the success of this wheel is not just in the ingredients, but, they report, in the "LEVICAST" method of manufacturing the wheel. You can download their paper from the Journal of Materials Science and Engineering A, Volume 329-331, June 2002, Pages 582-588, through Science Direct http://www.sciencedirect.com. In their paper they report on testing their new lightweight high-temperature TiAl alloy wheel in a Mitsubishi TD05 turbocharger on a Mitsubishi Lancer Evo. Being lighter than an Inconel wheel it spools up faster and has a significantly higher burst tip speed. After extensive testing this turbocharger was adopted by Mitsubishi for their WRC car.

 

[GeoffBob]

Shown below is the turbine side of the turbo from a Maxi5. The arrow points to the numbers "247" stamped on the end of the turbine wheel indicating that the turbine is made from the alloy Mar-M247, as opposed to the more standard Inconel. Mar-M247 or "Marchem" went on to be used in quite a few turbochargers fitted to racing engines. Note the A:R ratio of 0.63 (quite low) on the turbine housing. Even on a 1500cc engine this turbo would have been quite fast to spool up, which is of course exactly as was intended given that this was a rally car.

Most modern day production turbo cars employ a turbine with a low A:R value to minimise turbo lag (the time it takes to spool up the turbine before the onset of boost). I was told the other day that the 2.5l Ford Focus ST has a turbine A:R of 0.38!! I find this hard to believe, but if this is true it must lead to awfully high turbine temperatures if boost is sustained for any great period of time (remember, manifold pressure can be related back to turbine pressure), which could lead to to fatique cracking of turbine housing with time. I was told that this is exactly what happens, but that the average driver will floor the accelerator for only a short period of time before letting go and that the Focus ST wasn't built as a track racer (at least not in stock form). Well if all this is true I honestly don't know, but the turbo will apparantly out last the warranty.

 

[SJ Morgan]

Rather than get carried away with the DPV units, which by now are obsolete anyway (It's quarter century old tech), why not just contact your Garrett dealer and get a ceramic bearing replacement GT series turbo to replace the ancient T03? This gives the advantage of improved aerodynamics of the latest compressors and turbines, as well as a quicker spooling and longer lasting bearing to replace the old bushing.

Though total boost would not be increased (unless you want more boost), spool time would be greatly reduced, and throttle response off of boost would be far quicker. It's a simple installation, and you get 60 years of development for the cost of a new turbo. Simply put, I bet that a new designed turbo would be just as responsive on the road as one of the rare DPVs, cost less and offer no additional complexity.