john banks

4th gear WOT boost plot with only actuator tension (about 0.6 bar on car cracking pressure). Backed off because about to hit fuel cut.


No pictures, but the turbo housing had a splitter welded in.

Despite the wastegate penny being able to open to 45 degrees, when it does it basically blocks its outlet. This is graphically revealed with the boost creep when you have a turbo housing splitter in situ. So the downpipe pictured at the top basically appears to flow very little down the wastegate tube and when it does it is tortuous and restrictive.

I have the later version of the same downpipe at the top, and it does not boost creep with a splitter, but there is no divider in the downpipe flange. It still appears compromised though. Wonder if this is why I did not gain even 1 BHP from going to 3" system?

Wonder how the various twin dumps fare in this respect. Andy F has the Magnex twin dump and he put a splitter in and his 20G boost crept a lot. Mine does not with the same exhaust housing and actuator.

Boost creep is very dangerous, and can make you think your undersized turbo is great at holding top end boost, whereas really you are about to nuke the compressor wheel by sending blade tips supersonic. EGBP is ridiculously high in this situation. It is very bad for your engine, and an unsafe failure mode from a control point of view.

If putting a splitter in the turbo housing you'll soon find out that some of them actually are no better than the original design in that the wastegate gas can only shoot across the turbine outlet, which is obvious from the design of the original and the HKS downpipe, but not from the pictures of aftermarket twin dumps.



In summary I am saying that the downpipe pictured at the top is basically no better than the Subaru original at the bottom apart from being catless and 3".

[Edited by john banks - 4/29/2003 6:24:45 PM]

David_Wallis

is this why the wrc turbos have the penny pivoting differently?

David

jameswrx

iTrader: (4)

I switched to an open neck downpipe (with my own splitter mod) to try and elliminate boost creep...it didn't work. My car(94 wrx) with only hks exh and my d/p and air filter comes in on WOT at 0.85 bar.. it then boost creeps until standard fuel cut (just before rev limit)

The only way my car will not boost creep is if I run the scoobyecu module. When boost comes in at 1.15 bar it will hold this or drop marginally up to rev limit in 3rd. It's like the turbo wants to be running around 1.1bar and if it doesn't come in on WOT at 1.1bar(ie when standard) it tries to get there..lol

Question is why do these cars boost creep. Is it the engine design?. I say this because I've had 7 other turbo charged cars(not impreza's) and played around with them all boost wise
(even swapping exh turbine housings etc and I've never seen boost creep until the Impreza.

Is it anything to do with the distance of the turbo from the exhaust ports?

Pavlo

Whereas I would report the opposite for open neck DPs with no splitter in the neck.

Absolutely no boost creep on car with same turbo as Andy F but standard H&S DP.

Struggled to get boost holding on David Wallis' car with similar exhaust, but no creep

I struggle to get over 13psi on my Legacy with open neck exhaust, but other legacy owners can run 14.5psi no problem.


Paul

dowser

Isn't the real issue that the splitter should *not* protrude past the D/P flange? I remember reading that for max torque effect you jst have to stop the WG gases from hitting the Exhaust gases directly....boucing them off a small splitter is fine

I never understood twin dump anyway - without a splitter it's doing nothing (all gases diverted to main exhaust exit), with one you get boost creep.

Richard

P20SPD

iTrader: (1)

So a wastegate that opens vertically instead of horizontally would be better with a downpipe that has a splitter on the flange???

rex11

I have the BPM twin dump downpipe with a MBC and I don't have boost creep problems. I have the target boost set at .1 MPa and even under heavy loads (going up steep hills in a high gear) the boost stays at .1+/-.025 as read on the gauge. I'm running the stock MY01 TDO4, so the lack of top end puff could be masking the creep.

Some of the DSm guys have gone to a larger size wastegate flapper and ported the housing to allow for better opening characteristics. I don't know if this will help but try www.team rip.com for the parts.

nom

I think the two gas flows just need to be allowed to join properly - if the gas is moving fast enough & the angle is 'hard' enough to try to slam them together, they will do the same as any fluid & give an effective brick wall effect. So the gases from the wastegate don't have anywhere to go...
If the join is smooth & further down the DP (look at the BPM for how to do it properly!), theoretically the flow should be better than with the bellmouth (even with a splitter that's designed/made correctly) & have no wastegate-blocking problems at all. I think a lot of the 'bad press' that twin-dumps get is because most of them simply connect the two gas flow (if it ever even seperated them in the first place!) back together again as soon as possible - how can this do anything other than block the wastegate?

john banks

My understanding is that they boost creep because you have excessive restriction somewhere between the start of the wastegate tunnel and the full flow area of the downpipe.

I reckon the influencing factors are:

* small diameter wastegate tunnel
* long wastegate tunnel
* limited angle of wastegate opening
* splitter that gets a good seal
* twin dumps especially where the pipe is small, long and obstructed by the wastegate penny
* efficient compressor which pushes more air flow giving more for the wastegate to deal with
* efficient turbine wheel
* too small turbine wheel

It seems like a self generating run-away process, slightly too much for the wastegate to handle and the turbine speed increases, this speeds up the compressor and the thing decompensates quickly.

Boost creeping turbos (if you are able to hang on long enough to see) do seem to stabilise at a given level - presumably where the compressor efficiency is waning.

It is clear the downpipe has a huge influence - you can have no boost creep at all with one and horrendous boost creep on another with say a VF24 which is not known to be creepy.

The owner of the downpipe in the post has just removed the splitter from the turbine housing (unfortunately the downpipe does not look amenable to surgery) and found that the creep is reduced to 1.4 bar at 6000 RPM, dropping to 1.2 bar at 7000 RPM in 4th gear which will at least be mappable to 1.5 bar to 6000 RPM boost targets, but will not give as smooth control and accurate control as a turbo which you can just throw say 60% duty at the solenoid and it just reaches 1.5 bar nicely and then rolls off naturally after 6000 RPM. JECS at least likes to home in on a boost target and make small adjustments to keep it there rather than step from 60% duty cycle at 4000 to 20% at 5000 and 0% at 6000!

I think he should port the wastegate next or switch to another downpipe. What do you guys think?

I don't think the TD04 boost creeps because the compressor never makes enough airflow to overflow the wastegate which is more than adequate.

jameswrx I would be very interested to know if you ever remove the splitter from your bell mouth downpipe if you cure the creep. I have a sneaky suspicion it would probably be enough to do so.

BTW, the same car with the same spec with a different turbo with a ported wastegate and actuator at least twice as strong just would not hold top end boost, and with the pneumatic feed to the actuator completely disconnected only made 1.35 bar midrange!

It seems we need a careful matching of turbo to splitter and downpipe to avoid the extremes.

Obviously Subaru have it right from the factory - if you throw a constant duty cycle at a bog standard setup it reaches a nice target with no real spike, holds it and then magically rolls it off in a sensible fashion to suit the size of the turbo.

But when we modify it and change the system it needs a bit more work to get OEM levels of control back

[Edited by john banks - 4/29/2003 9:04:52 PM]

jameswrx

iTrader: (4)

surely if you can get the wastegate to open 90 degrees it's not going to matter which way it opens?.

Still something more in boost creep for me than basic downpipe design. Std, hks, open neck all creep on my car, but with ecu module in place seems fine. Use a dawes device and fuel cut raiser giving target boost of 1 bar coming in..boost creep.

Better solenoid control with scoobyecu module?

How does the solenoid control boost levels?, does it open and close quickly?, or bleed off with varying opening?

john banks

James I think you are just masking the boost creep by running higher boost than the creep level. This is a valid approach but not good as a failsafe. If your ECU decided to run zero duty cycle because of horrendous knock it would not have any control for example.

When testing for boost creep here I think the ECU's boost control method and what solenoid you are using are not relevant - I prefer to directly connect the compressor outlet to the actuator.

It is VERY interesting that your standard downpipe boost creeps though! Maybe removing your splitter wouldn't help then! Was this with standard mid section and backbox?

Most Subarus have no boost creep issues at all from factory AFAIK.

The new age STis are most famous for it with a turbo back exhaust, but they don't boost creep as standard.

If you had vertical opening, would the wastegate flow not just hit another area of the pipe and bounce about?

jameswrx

iTrader: (4)

last time I tried stad d/p was with hks mid and back.

The car was completely std when I got it so cab't be certain it crept 100% std, but I'm sure it did.

I only had the car a week before putting exh and filter on and had never experienced boost creep before (previous cars). I do remember the car speeding up at the top of the revs though and obviously you get this feeling when it creeps. I never actually studied the gauge at this time, I did get a recorded maximum of ~0.90bar though.

Also intersetingly when I done the exh change I checked the preload on the actuator and found it was actually turned out so the wastegate door was actually open??. This could mean someone tried to dial it out before?.

nom

Just sneaking back to my flow stuff again ...
Think of a fast jet of water flowing freely. Now spray a second, smaller jet parallel to it & merge the two. All should go nicely. Up the pressure on the main flow, and they'll continue to merge well for some time. Next, have the same main jet of water flowing, and spray a jet in from the side (60 deg or so to the main flow - the 60 deg making the flow closer to perpendicular than parallel, that is). It'll make a mess, as the bulk of the flow from the secondary pipe will go through (think about this as well in a restrained system...). Now up the pressure of the main jet - the effect is that the smaller side-jet will start to bounce off. Now, add that restraint again... where does the flow bounce off to? Oh yes, back where it came from, as there's nowhere else to go... So, hello something trying to shut the wastegate again.
This is likely to happen with badly designed twin dump, and much less so for bellmouth, as the flows just get into a jumbled mess rather than being directed, upping EGBP in general rather than directing it largely at the wastegate.
Most of this may well be pants of course, but one things I have usual been able to do is fluid dynamics (until it goes supersonic, that is, which might be a little scary in this case )
So. In answer to the question as to what to do, it's change the downpipe. I can reccomend a particularly nice example that Christian is still trying to sell, I think...

rex555

I installed a TDO5 16G in my WRX recently and had boost creep problem using my existing Bosal twin dump dp. I connected the turbo directly to the acutuator and the sam eboost creep occured. I changed to a wide mouth dp and boost creep is gone. It really appears that the a horizontal opening wg on a high flowing turbo does not like a dp with a spliter on it.

blachie

i had boost creep with the std td04l+3in closed neck trust dp and std ecu boost control,in 5th wot and 4000rpm i used to get 1.25barfor a sec or two settling to 0.9-0.95bar.i "open-necked" it-it looked exactly like the h&s one- and the problem remained unchanged.
i then changed to an aps high flow turbine outlet with a spliter mached for the td04l,and boost creep was reduced to aproxx 1.1bar(0.95target)i then added a dawes set at 1.05bar(1.5mm hole) on a cold night and everything is perfect now.
IMHO the design of the dp in the first picture is poor to say the least.
1.flow from the turbine and wastegate discharge is at 90deg creating turbulence-backpressure.
2.it looks to me that if and when the wg door is open it blocks completely the small pipe.
3.when the wg door is closed some of the exhaust gases could be "recirculating" backwards from the small pipe back and at 90degrees to the main turbine discharge ,a correct length spliter would do wonders with that dp.
all of the above is just food for thought...have a look at the aps site to see the aps hfto,i think the aussies have the perfect design in their hands...

[Edited by blachie - 4/30/2003 2:05:37 PM]

Denmark

iTrader: (2)

Blachie,
Thats not boost creep,

It´s boost´creep when the boost rises in high rpm without bieng able to lower it

Skassa

P20SPD

iTrader: (1)

Thats a boost spike aint it?

Denmark

iTrader: (2)

Yup

harvey

iTrader: (48)

Big open mouth on the downpipe and no splitter.

GDBSTi

I have tried a twin dump, bellmouth and splitters.No luck. Only way I have solved boost creep on VF35 and VF 23 turbo completely is to take the turbo off and port the path leading to the wastegate (inside the exhaust housing of the turbo) for the exhaust gases to flow easier through the wastegate instead of through the turbo. Even small imperfections in the castings ( not all VF35's have boost creep with catless 3inch exhausts, mine did and had small imperfections in the casting around the entrance to the wastegate from the exhaust side!!!) or even carbon build up(friend with a well used VF23 on 1999 Turbo running rich for safety removed carbon build up and problem solved) seem to effect the ability of the wastegate to divert exhaust gases. It would be interesting to know if EVO's also get boost creep??