Ford Focus ST Forum banner

81 - 98 of 98 Posts

·
Administrator
Joined
·
11,294 Posts
Yikes I just bought an ATP turbo kit. How common is this failure? I’ve never even heard of it until now.


Sent from my iPhone using Tapatalk
It depends. LOTS of people run the ATP kits, and the failure rate isn't that high. It is a design error. Their entire elbow EWG setup doesnt flow great.

My next turbo kit is CP-e for sure.

Mine has been good so far, I forget how many miles. It's been 3 years though iirc.
 

·
Premium Member
2013 ST2, AGP BWS turbo, SP63 block
Joined
·
478 Posts
Discussion Starter #82 (Edited)
Yikes I just bought an ATP turbo kit. How common is this failure? I’ve never even heard of it until now.


Sent from my iPhone using Tapatalk
Well I should say, the mode of failure I just laid out is speculation. I will be doing a CFD analysis on this for my class next semester because I think its quite possible that the oversized design is the leading problem.

For my Penn state senior project I did study thermal fatigue failure due to very similar reversing edge pipe flow in the header of a massive heat exchanger ( for a heat exchanger in a HRSG of a power plant) And found that pipe junction corners did indeed promote thermal cracking.

Here's a sketch of the Fluid problem. Notice the reversing flow over the edge that the ATP design creates. Exhaust gases are extremely hot and fast and have high pressure. They are full of energy and generally where they swirl, it creates widely varied heat and stress profiles in the solid material that is housing them. (not drawn here. just the flow is drawn.) This varied heat and pressure profile through the solid material eventually leads to cracking. Uniform heat/pressure distributions through solids are much easier on them (especially aluminum)
390123




Here's a screenshot from Valley Crew racings teardown. you can see the crack and the oversized ATP ring footprint.
390122

Here's the video where he mentions the failure at the beginning


But the fact that @slo.shoten just had this happen on a stock turbo may point more to a bad design on ford's behalf.
Or did the previous owner of @slo.shoten's car did have an ATP kit on it? idk.
This is why I want to research it.
 
  • Like
Reactions: junits15

·
Registered
15 TB Focus ST3
Joined
·
447 Posts
It depends. LOTS of people run the ATP kits, and the failure rate isn't that high. It is a design error. Their entire elbow EWG setup doesnt flow great.

My next turbo kit is CP-e for sure.

Mine has been good so far, I forget how many miles. It's been 3 years though iirc.
I see they have a Precision Flange now for their atmosphere downpipe kit. Me wonders if any of their flanges will fit a G25-660, because I love the kit.
 

·
Registered
Joined
·
27 Posts
But the fact that @slo.shoten just had this happen on a stock turbo may point more to a bad design on ford's behalf.
Or did the previous owner of @slo.shoten's car did have an ATP kit on it? idk.
This is why I want to research it.
I'm the only owner of the car and I have never had an ATP turbo on it. The car has been mostly stock for the majority of it's life. The only "major" mod I did about 10,000 miles ago was installing a COBB accessport with a tune from Mountune.

I'll try to post some pics later, but I can't see the cracks on the headifold outlet, the cracks were deeper inside the ports according to the machine shop. I'll see if I can get a borescope or something in there to check it out.

Personally, I think this headifold design is awful. A traditional 4 port head and manifold would have been much better. The machine shop has seen this same failure (cracks in the exhaust port to a water jacket) on Honda heads that use the same "headifold" design.

A lot of heat from the exhaust transfers into the coolant because of this head design also...
 

·
Premium Member
2013 ST2, AGP BWS turbo, SP63 block
Joined
·
478 Posts
Discussion Starter #85
Also how did that embiggened stock downpipe do? Idk if you ever mentioned that in your other threads lol


Sent from my iPhone using Tapatalk
I did follow it up somewhere deep in that "replace cartridge (Big wheel stocker)" Thread. but I didn't do a full review on anything in my build yet for lack of time and lack of quantitative dyno data.

It does sound a little bit throatier/louder, but just barely. Does not increase overall db in the cabin on the highway.
The best part about it is that even without a (stupid) crackle tune, It will occasionally burble like a high flow downpipe. I've long since quit screwing with the crackles experiment. I'd like to keep this cat for a while and it burbles politely on its own now.
It is fine with a stage 3 tune, I wasn't seeing any issue with OTS ignition timing or Flange temps or Back pressures there.
But it's tough to say if that would be the same for everyone. My turbine blades are clipped.

As far as gains, I have no quantiative evidence. And I'm going to be tuning the VCT for these mustang cams once the motor is put back together so by the time it sees a dyno, there will be no way of knowing where the gains are coming from.
 

·
Premium Member
2013 ST2, AGP BWS turbo, SP63 block
Joined
·
478 Posts
Discussion Starter #86
I'm the only owner of the car and I have never had an ATP turbo on it. The car has been mostly stock for the majority of it's life. The only "major" mod I did about 10,000 miles ago was installing a COBB accessport with a tune from Mountune.

I'll try to post some pics later, but I can't see the cracks on the headifold outlet, the cracks were deeper inside the ports according to the machine shop. I'll see if I can get a borescope or something in there to check it out.

Personally, I think this headifold design is awful. A traditional 4 port head and manifold would have been much better. The machine shop has seen this same failure (cracks in the exhaust port to a water jacket) on Honda heads that use the same "headifold" design.

A lot of heat from the exhaust transfers into the coolant because of this head design also...
Interesting.. good to know! Best of luck on finding a new head. They are tough to come by.

Yeah it definitely heats the coolant right up haha.
I see why ford wanted this design. The less volume between the turbo and the piston, the better. Additionally, cooling the manifold may have some mysterious benefit too.. so long as it didn't become a failure point (shame).

They just didn't spend enough time stress testing this thing. But the failure modes are difficult to analyze. Fluid flow is an actively evolving field. we haven't even had the computer strength to solve these types of problems until recently.

I wonder if the 2.3L exhaust headifold cracks as much. They definitely dropped the ball on the head gasket side of things.
 

·
Premium Member
2013 ST2, AGP BWS turbo, SP63 block
Joined
·
478 Posts
Discussion Starter #87
I'm the only owner of the car and I have never had an ATP turbo on it. The car has been mostly stock for the majority of it's life. The only "major" mod I did about 10,000 miles ago was installing a COBB accessport with a tune from Mountune.
How many miles do you have on it?
 

·
Registered
Joined
·
27 Posts
How many miles do you have on it?
I drive the car a lot, it's my daily. It has 96,XXX miles on it. This is the first-ish major mechanical problem with the car. The only other thing I had fail, recently, was the dual mass flywheel.

I can see some benefits to the head design. This design would help the coolant to reach operating temperature very quickly. Getting the coolant up to temp quickly would also reduce startup wear and help lower emissions on warmup.

Not sure if this is accurate, so take it for what it is, the machine shop mentioned that cooling the exhaust manifold would help prolong the life of the catalytic converter....not sure if I trust that theory though.

Fortunately, the dealer was able to order a cylinder head. I hope it's not delayed though...I don't really feel like scrounging the junkyard to harvest a head off another Focus.

Speaking of fluid flow, I wonder if extrusion honing the exhaust passage would improve flow and reduce the chance of turbulance/hotspots...

It would also knock down any sharp edges and protrusions that could get hot and cause heat stress/fracture points.
 

·
Administrator
Joined
·
11,294 Posts
I drive the car a lot, it's my daily. It has 96,XXX miles on it. This is the first-ish major mechanical problem with the car. The only other thing I had fail, recently, was the dual mass flywheel.

I can see some benefits to the head design. This design would help the coolant to reach operating temperature very quickly. Getting the coolant up to temp quickly would also reduce startup wear and help lower emissions on warmup.

Not sure if this is accurate, so take it for what it is, the machine shop mentioned that cooling the exhaust manifold would help prolong the life of the catalytic converter....not sure if I trust that theory though.

Fortunately, the dealer was able to order a cylinder head. I hope it's not delayed though...I don't really feel like scrounging the junkyard to harvest a head off another Focus.

Speaking of fluid flow, I wonder if extrusion honing the exhaust passage would improve flow and reduce the chance of turbulance/hotspots...

It would also knock down any sharp edges and protrusions that could get hot and cause heat stress/fracture points.


People have extrude honed the headifold, but the build was like 500 hp, and the car got parted, I forget what turbo he was running. It will help for sure.

Most people just do valve work, and leave the exhaust port alone. That water jacket is pretty close.

Im not sure what our whp record for the single exit head is. The 2.3 head has bigger ex valves. So it flows better, and then the twin scroll mates up better with with circular headifold adapters.

Theres a lot going on here. A lot that may be excessive thinking.

Our platform does fine into the 550whp ish zone. So unless you are going to build something big, most of this is a bit much. The 2.3 head has done us well into the 650-700whp realm.


Longevity being the concern, thats tough to get results on. Most people pop their stuff, crash, or sell.

Not sure if theres many high mileage big builds kicking around. Could maybe comb through the build threads and find somone, but finding an adequate sample size may be tough.
 

·
Registered
Joined
·
27 Posts
Longevity being the concern, thats tough to get results on. Most people pop their stuff, crash, or sell.

Not sure if theres many high mileage big builds kicking around. Could maybe comb through the build threads and find somone, but finding an adequate sample size may be tough.
Longevity is my main concern, of course, I do want a bit more power and fun drivability from the Focus, but I would like it to last a long time also. Personally, I don't think 200k is asking too much of a modern engine. Call me crazy :ROFLMAO:
 

·
Moderator
Joined
·
2,745 Posts
@Habriant40 Looks like the BNR and EFR elbows would fix this issue, as the hole matches the head port. So even if theres an increase in diameter beyond that. The swirl would happen against the back side of the iron flange and not the cylinder head.
 
  • Like
Reactions: junits15

·
Premium Member
Joined
·
1,248 Posts
It depends. LOTS of people run the ATP kits, and the failure rate isn't that high. It is a design error. Their entire elbow EWG setup doesnt flow great.

My next turbo kit is CP-e for sure.

Mine has been good so far, I forget how many miles. It's been 3 years though iirc.
Yeah but you saw me do that transmission lol my luck indicates that I’ll blow my head in 500 miles ;)

Plus there could be power on the table here, that port mismatch is not ideal and could be impeding flow rate.


Well I should say, the mode of failure I just laid out is speculation. I will be doing a CFD analysis on this for my class next semester because I think its quite possible that the oversized design is the leading problem.

For my Penn state senior project I did study thermal fatigue failure due to very similar reversing edge pipe flow in the header of a massive heat exchanger ( for a heat exchanger in a HRSG of a power plant) And found that pipe junction corners did indeed promote thermal cracking.

Here's a sketch of the Fluid problem. Notice the reversing flow over the edge that the ATP design creates. Exhaust gases are extremely hot and fast and have high pressure. They are full of energy and generally where they swirl, it creates widely varied heat and stress profiles in the solid material that is housing them. (not drawn here. just the flow is drawn.) This varied heat and pressure profile through the solid material eventually leads to cracking. Uniform heat/pressure distributions through solids are much easier on them (especially aluminum)
View attachment 390123



Here's a screenshot from Valley Crew racings teardown. you can see the crack and the oversized ATP ring footprint.
View attachment 390122
Here's the video where he mentions the failure at the beginning


But the fact that @slo.shoten just had this happen on a stock turbo may point more to a bad design on ford's behalf.
Or did the previous owner of @slo.shoten's car did have an ATP kit on it? idk.
This is why I want to research it.
You make a good point we really don’t have any solid evidence linking this failure to the ATP flange other than anecdotes.

I’d really like to see what your simulation indicates though, it’s rare that we see any engineering out in the open like that. I feel most of it it happening behind closed doors.

Just as long you have enough system memory for ansys

Does anyone make a garret compatible elbow? A lot of people buy ATP kits a better flowing elbow would be a good seller. And for purely selfish reasons, if I can buy a better elbow to put on before I tear my car apart that would be nice lol.

The CPE kit is precision or tial compatible depending on which one you get.


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
27 Posts
390168


Well...the crack is there and pretty plain to see. It's not cracked on the edges of the outlet, but rather inside close to that divider.

Maybe I just had a "bad casting" that lasted 96,000 miles? I do drive the car in a spirited fashion but I don't beat on it. Also I've never taken it to the track. So perhaps that's why it lasted so long before cracking.
 

·
Premium Member
2013 ST2, AGP BWS turbo, SP63 block
Joined
·
478 Posts
Discussion Starter #94
Well...the crack is there and pretty plain to see. It's not cracked on the edges of the outlet, but rather inside close to that divider.

Maybe I just had a "bad casting" that lasted 96,000 miles? I do drive the car in a spirited fashion but I don't beat on it. Also I've never taken it to the track. So perhaps that's why it lasted so long before cracking.
Thats wiggity wiggity wack.
A little more wall thickness could have really helped them out.

it’s rare that we see any engineering out in the open like that. I feel most of it it happening behind closed doors.

Just as long you have enough system memory for ansys
I think simulating real world applications (both numerically or analytical by hand) take so much time to to plan, understand, model, and validate that you usually only see them done for pay, even with the luxurious software suites like ANSYS.
In a similar sense, a hobby welder could weld up a roll cage any time they feel like it, but 90% of roll cages in existence were probably paid for.

haha I really could use more memory at home.. Its nice running them on my work's computer (64gb DDR4 ram).
I still have to write up something about the Fluent simulation I did for the charge pipe baffles.
 

·
Registered
Joined
·
27 Posts
Depending on where you are located I could build you a PC 😁. I'm rocking "only" 32GB of ddr4 at 3600mhz. i9-9900k, rtx 3070, and 6TB of Enterprise Intel SSDs. I work in IT and was able to piece this build together over time. I imagine it would run simulations pretty well....but I have never even heard of ANSYS until now.
 

·
Premium Member
2013 ST2, AGP BWS turbo, SP63 block
Joined
·
478 Posts
Discussion Starter #96
Depending on where you are located I could build you a PC 😁. I'm rocking "only" 32GB of ddr4 at 3600mhz. i9-9900k, rtx 3070, and 6TB of Enterprise Intel SSDs. I work in IT and was able to piece this build together over time. I imagine it would run simulations pretty well....but I have never even heard of ANSYS until now.
That sounds like a beast of a simulation machine. SSD makes it even easier because if do use up the ram, the time it takes for ANSYS to offload into storage is minimal. I'll be building one eventually here.. Just as soon as my ST stops suckling my $$$.

ANSYS is a simulation suite that was made by a bunch of absolutely brilliant Engineers. It takes the crazy greek math that describes how nature works and solves it. The way you can solve most of those problems by hand is by simplifying to 1D or 2D and making a bunch of assumptions. Most complicated problems that require 3D solutions can not be solved by hand.

It takes a 3D cad model, you tell it about the physics and conditions of the 3D cad model, It turns that info into complex array of equations/inputs and crunches it with iterative solving. I have filled the entirety of my 64 gb of ram with active solver data. The amount of data it generates is huge. A solution file can be like 8gb for a simple charge pipe (wasn't even the whole charge pipe, just the baffle region.)
 
  • Like
Reactions: junits15

·
Premium Member
Joined
·
1,248 Posts
Thats wiggity wiggity wack.
A little more wall thickness could have really helped them out.



I think simulating real world applications (both numerically or analytical by hand) take so much time to to plan, understand, model, and validate that you usually only see them done for pay, even with the luxurious software suites like ANSYS.
In a similar sense, a hobby welder could weld up a roll cage any time they feel like it, but 90% of roll cages in existence were probably paid for.

haha I really could use more memory at home.. Its nice running them on my work's computer (64gb DDR4 ram).
I still have to write up something about the Fluent simulation I did for the charge pipe baffles.
Finding people who have the right background, have access to the right tools, have the right car, and have a willingness to apply their background and tools to issues with the car is a rarity.

Most guys just wanna leave work at work.

I’m an RF designer so I use hfss at work, I’m currently pegging all 32 cores at 100% and using 50gigs of ram on my workstation with a simulation lol. HFSS uses the same finite element modeling technique applied to maxwells equations through a 3D model.

So unfortunately I can’t be of any help with what you’re doing, and I can’t really contribute too much of my abilities to this kind of stuff, but I know enough about finite element modeling to be able to see when it’s applied to a problem

Either way I really like seeing what you’re doing, I get so tired of seeing parts for this car that I can intuitively tell are poorly designed but have no background or skills to prove.

From what I see there are some people out there who actually take the time to find well engineered solutions for issues with this car, I think CPE and mountune are my favorite two vendors for that reason. There are also people who just throw large pipes and flashy colors at the issues, without considering what that means for actual performance.
 

·
Premium Member
Joined
·
1,248 Posts
FYI
For anyone who may be interested

Looks like the full race elbow can be configured to work with a Garett turbo upon request. No idea how much that would be but I assume it won’t be cheap as it’s custom work.




Sent from my iPhone using Tapatalk
 
81 - 98 of 98 Posts
Top