Understanding Turbo Compressor Maps

[Guy Chiattello]

Over the past several months I have been reading books and studying information on turbo's. I have a question. At lower RPM the compressor is naturally going to be in the "surge: range. Does this cause problems with engine performance? If not what is the real significance of "surge"? I have read all of the popular books and no one seems to really explain this.

Guy Chiattello
82 Black coupe

[Hladun]

Surge is like stall on an airplane wing. The turbine blades lose effectiveness (lift). The result is damaging to the compressor. Flow, not rpm, is the problem. Description of surge in turbine engines is probably the best source of info.

[tsuter]

When you enter the map way to the left on or near the surge line you have low compressor efficiency so you get slow spool up as well as premature bearing failure. On the other hand it means as the RPMs come up you enter the high efficiency area and have excellent top end performance.

If you choose a map where you avoid the surge line and enter at a higher efficiency then you get better spool up sooner but at the higher RPM you risk the overreving the turbo.

Most the maps end up not being as flat as desired so you need to choose based on low end or high end or compromise.

Here is a very good site:

http://www.turbofast.com.au/turbomap.html

[sammyg2]

Surge is described as a shockwave that travels back and forth from suction to discharge in a compressor. It happens when the compressor is unable to compress the gas (or air in this case) enough to maintain flow. If there is 8 psig backpressure in the compressor discharge piping and the compressor is only capable of producing 7.5 psig at a certain speed, the flow will stop. Then the higher 8 psig will reverse flow towards suction. As it does this the discharge pressure drops until the compressor can again produce enough discharge pressure to re-establish flow in a positive direction. This surging can happen very quickly, as many as 270 times a second on certain machines I'm familiar with. Another by-product of surge is heat. If you compress a gas it gains heat. during surge conditions that hot gas can reverse flow into the suction where it is compressed again and gets hotter. If allowed to continue the temperature can continue to rise until it gets to damaging levels.
the shock wave can also cause damage because of the radical changes in loading that happens. The impeller goes from full load positive to completely unloaded to reverse load and back to full load positive.
In some situations the surging can induce undersirable harmonics that are amplified by natural resonant frequency of the rotor and rpm. I've seen a 14 pound compressor rotor come apart at 52,000 rpm due to this "critical speed". not pretty.
The severity of the damage is directly related to the severity of the surge. in other words at low pressure and flow the surge is not severe and probably won't reduce the life of the machine noticeably. At higher pressures the surge is more pronounced and severe and can cause damage very quickly.

You'll notice on most compressor maps (or curves) the surge line is not continued all the way to the right or zero line, they usually stop at around the 20% of rated flow or the minimum flow line. To the left of that line is not desirable, but not that severe for short periods of time.

The farther you get to the right of the map the more severe the surge is likely to be if the pressure rises high enough to stall the compressor flow.

Small turbochargers are pretty tough when compared to jet engines or large industrial centrifugal compressors. A smaller turbo can take quite a bit of surging where much larger machines cannot.

[Guy Chiattello]

Thanks for the input everyone. Tustuer, I understand your comments about finding a compromise between low end and high end performance.
However, I am considering a twin turbo conversion, with turbos' installed at the ends of my SSI's and then going into intercooler in each of the rear fenders. I have installed a C2 tail so have little room for an intercooler under the deck lid. Also, I would plan to replace the CSI with Tony's Megesquirt kit. Looking for 7 to 9 PSI boost.

According to the compressor maps, With two smaller turbo's (Garret t3 with 40 trim), I should be able to experience boost at about 2K rpm (about 4 lbs/min per turbo) and about 10 psi at 6K rpm, or about 14 lbs/min per turbo. With the turbo's being in surge below 2K rpm, will this be problem? I wouldn't think so.

A further question is, by having twin turbo's do I expand the boosted torque curve, over a single turbo? What are some reasons I may not want to go with twin turbo's?

Here is a neat example of rear mounted twin turbo's on 2.1 liter engine which produces in excess of 400 HP.

http://www.patwilliamsracing.com/racedrivetrainphoto2.1tt.htm

Guy Chiattello
82 SC Black Coupe

[beepbeep]

Two T3's are tad too big for 3.2L displacement IMHO. I run one T3 - 50 on 16 valve 2L engine and boost threshold was around 2500 RPM.

WHat you are trying to do is to bolt T3 on what is basicaly 1.6L engine per turbo. It's going to be good for top-end power (as they can flow 400hp worth of air together), but that advantage is lost in your case beacuse you are keeping your OEM C/R and using low boost.

I recomend two T25's or KKK K03's...they will spin up faster.

[tsuter]

I agree with the K03. They are designed for the VW1.8T and deliver very quick spool up. Two of these on a 3.x would be very interesting!!.
I have swapped in a K04 on my 1.8T and it delivers about 260TQ and 235HP. Its is spooling at 2000RPM! So the sky is the limit.

As beepbeep says keep your CR as low as possible NA (8.5 or lower)even if you have to change pistons then run the boost up to 10-12.

I have an extra K03 you can have if you want to pick it up in Madison.

The new Garrett GTs are very nice but the MAP requires much higher boost than you will be able to run to really get into their efficiency islands. I've looked at a ton of them for my 3.0 and can't find anything that is efficient in the 3-10 lb area before about 4000RPM.

Twins would definitely solve that.

P.S. if you ever get to Madison I can introduce you to some of the guys at Kelly Moss (kellymoss.com). They do some interesting things.

[Guy Chiattello]

Thanks for the T25 or K03 recomendations. I did look at the map for the Ko3 and saw that it was a possibility but didn't know how to read the X Axis Tsuter I do have business in Milwaukee on a regular basis so will try to make an effort to get with you. I will e-mail you early next year.

Guy Chiattello

[beepbeep]

There are so many parameters when deciding which turbo to bolt that it's always a mix of calculations and empirical knowledge.

Unfortunately, most of new small turbos have one thing you cannot provide: watercooling

[tsuter]

The X-Axis should be air flow. Here is the theoretical flow for a 3.0 at RPM.
Theoretical Air flow lbs/min:
2500 = 19
3000 = 23
3500 = 26
4000 = 30
4500 = 34
5000 = 37
5500 = 41
6000 = 45

On a twin it would be 50% of this multiplied by the VE. (Volumetric efficiency) of the motor (70-100%) if known or 80% if not.

[tsuter]

The "beeper" is right about water cooling. These turbos (K03,K04) for the sports compact 1.6,1.8, 2.2 turbos etc. are oil lubed and WATER cooled.
If I didn't have a six speed project underway for the spring here is what I would consider:

Two K03s off each SSI outlet WITH water cooling! I would install a front spoiler mounted radiator and run it using the VW Afterrun coolant pump used by VW to water cool the K03. The afterrun coolant pump is small - about the size of the SC fuel filter. And you are not pumping much water so no need for a real waterpump.

That would be some trick set up!
Here is a picture to see how small it is.
If you don't have AC then put the radiator on the rear deck.

[Guy Chiattello]

tsuter,

I didn't realize the KO3's were water cooled. I am assuming you would still have oil feed lines as you would have with any turbo??

Your suggestion about using a small water supply and pump is very interesting. As I mentioned earlier I have a C2 electric deck lid and have modified my AC system with an Underbelly condenser. I therefore might consider putting a small radiator in the engine bay. I am considering running the intercoolers in the rear wheel wells. But that is whole new topic for discussion.

Would I gain anything with the watercooled turbo's? In other words, does watercooling give you anything beyond greater reliabiliy?

Guy Chiattello

[beepbeep]

Watercooling just gives you greater reliability, no other advantages. I've seen people running watercooled turbos without any water-cooling, and it actually worked...but they were very cautious with cool.down time before they switched off the motor.

There is a third posibility, and that is passive watercooling using convection. With other words no pump, just hot water circulating by itself.

[tsuter]

VW pumps water as coolant to the KO3 all the time. The afterrun coolant pump is on a thermostat and timer so it will also run as long as 5-6 minutes after the car has been shut down continuing to circulate water thru the K03.

The turbo is oil lubed so after a hot run you should still idle a bit to avoid coking the oil although this is pretty difficult to do.
There are both water and oil fittings on the turbo and it comes with an internal wastegate ie,. it is not free floating.

I have seen the internal wastgate welded closed and an external wastegate used. This is common as Tialco or other external wastegate is much smoother opening.

Attached KO3 picture shows the rough dimension of the turbine inlet from exhaust. You can also reclock the turbo (Turbine to compressor) housing for your fitment preference.

[Guy Chiattello]

tsuter,

Thanks for the picture. I believe the output of my SSC's are also
1 5/8 inches so there would be a good match up. Will however need to modify the flanges to the three bolt flange. Also gald to hear you could clock the compressor which you have to do, having one on each side.

I will get with you in January and you can let me know what you want for the k03.

Thanks for flow numbers as well. I will look at the compressor maps in light of these.

Guy Chiattello

[ohecht]

Guy,

Any update on your TT project? I am considering the same for my 82, and I knew there had to be others out there trying something similar.

Olivier

[ohecht]

Well,

I am starting the long process of trying to mount (first) and run (later) 2 new T-25s on my 82 SC. I got these because they were cheap, brand new, and at least I know they are in identical condition as I start.



I will be trying to make a coolant system for the bearings as part of the overall challenge, and hopefully develop an afterrun cooling function as well, since my wife is not good at things like cooling down turbos! I will also probably use an air-to-water intercooler since I have AC and will have a coolant system anyway (although I will probably keep the systems separate to avoid heat-soaking the coolant early and off boost). I also considered wheel well ICs, but there is no way I could fit on on the passenger side and I would worry about their effectiveness so close to the engine anyway.

This whole thing is really just an expensive excuse to learn to weld a little!

I have a "clocking" question for the turbo experts, though:

I can easily clock the turbine sides by loosening a few bolts. However, I need to also rotate the compressor housing around the center assembly in order to keep the wastegate actuators aligned. There is only a giant snap-ring on the compressor side. Can anyone confirm that just removing this ring will allow me to rotate both sides around the center housing? I don't want to mess up brand new turbos by taking the wrong thing apart.

Thanks,

Olivier

[Quicksilver]

On the subject of water cooled turbos...

What about cooling them with oil? A reasonably small sized restrictor should keep it from effecting the oil pressure (besides in a motor like this you would have had it opened up so it would have a larger oil pump, right?!).
The addition of an after run oil pump or accumulatot would finish the job.

I can't see a reason why oil wouldn't work well. Any thoughts, or has anyone tried this?


Wayne

[JP911]

I would think that the difference in viscosity between oil and coolant could cause problems. The cooling lines (at least on the 1.8T's KO3s) are pretty narrow and you could end up with coking problems leading to catastrophic turbo failure. I believe that the oil feed lines for the KO3 are larger in diameter than the coolant lines.

[ohecht]

Wayne,

I thought about the same thing, but I would also worry about how the oil would act inside the "water jacket" in the turbo, which I expect has very small passages for the more easily pumped water.

Olivier