Be Careful How You Box Steel 944 A-Arms!
10-02-2006, 02:21 PM
#46
Race Director
I believe they crack due to braking loads. When you brake you effectivly push the wheels back and chassis forward. Ball joint ends are actually rather strong. However at the bushing you are trying pry the bushing out in the plane of the arm. Imagine taking the bushing out and jamming a crowbar and tehn the plane of the arm wreching that crow bar back tring to "oval" the bushing hole. This creates intense stress on the forward point where this barrel meets the rest of the arm. The problem is that this is also where the steel makes a tight bend. This the problem. All these loads are occuring in the plane of the arm. Too bad I don't have a Nice clear picture of what I am talking about.
10-02-2006, 02:30 PM
#47
Race Director
Here is a picture showing what I am taking about.
Now the crack forms on the top side of the arm or the bottom side depending on which side the arm is on. If you look at the arm it is because the bend is tighter on one side and since the arm the same for both sides this tighter bend is up on one side and down on the other.
Now the crack forms on the top side of the arm or the bottom side depending on which side the arm is on. If you look at the arm it is because the bend is tighter on one side and since the arm the same for both sides this tighter bend is up on one side and down on the other.
10-02-2006, 02:31 PM
#48
Addict
Rennlist Member
Rennlist Member
Join Date: Jun 2001
Location: Claremont,NH
Posts: 586
Likes: 0
Received 0 Likes
on
0 Posts
I have a clear picture of what you are saying. I have worked with metal for 30 years. I have a pretty good idea how it all works and looks whan stressed.
you are right that all the streswss on a control arm is just what you are saying that is what I was saying is lateral stress.
But when you look at the pictures that TDin Dc sent the stress was in the other direction.
you are right that all the streswss on a control arm is just what you are saying that is what I was saying is lateral stress.
But when you look at the pictures that TDin Dc sent the stress was in the other direction.
10-02-2006, 02:37 PM
#50
Race Director
Without looking at the facture surfaces in detail it looks to me like the cracks I have seen are the start of TD's problem. Then once the crack got to certain point all hell broke loose and the thing broke off.
Hard for me to say for sure, but that would be my guess as I would not expect the arm to fail for any other reason , short of crash damage, there.
BTW... I once went off track and bent my control arm. The arm bent back like it would from extreme braking forces. Interestingly the arm did not fail at the barrel, but the rest of the arm bent and folded. Seems to me that breaking the barrel due to overload is less likely that failing it due to fatigue. Of course the arm that folded on my car was stock unreinforced arm. Had that been TD's arm it may have bent the frame or ripped the barrel off. Good news for me is that replacing the arm much cheaper than trying to bend the frame back.
Hard for me to say for sure, but that would be my guess as I would not expect the arm to fail for any other reason , short of crash damage, there.
BTW... I once went off track and bent my control arm. The arm bent back like it would from extreme braking forces. Interestingly the arm did not fail at the barrel, but the rest of the arm bent and folded. Seems to me that breaking the barrel due to overload is less likely that failing it due to fatigue. Of course the arm that folded on my car was stock unreinforced arm. Had that been TD's arm it may have bent the frame or ripped the barrel off. Good news for me is that replacing the arm much cheaper than trying to bend the frame back.
10-02-2006, 08:50 PM
#52
Addict
Rennlist Member
Rennlist Member
Hey;
It has always been my opinion that braking force was the main culprit in any bushing/arm load. I'm sure banging curbs doesn't help. Offs probably don't either. Anyone tieing the car on the trailer by using the wheels might be adding to this as well!
As you said, Bob, arm movement is not that great, and the rubber will certainly take care of that much deflection in a radial sense. The bushing material is pretty elastic and is made to have the sleeve trapped between the x-member flanges.
A friend had his L/F snap off as he set up for that nasty off camber right into Thunder Valley at Mid Ohio. He went straight off and the gravel trap stopped him about 3-4' from the wall.
It has always been my opinion that braking force was the main culprit in any bushing/arm load. I'm sure banging curbs doesn't help. Offs probably don't either. Anyone tieing the car on the trailer by using the wheels might be adding to this as well!
As you said, Bob, arm movement is not that great, and the rubber will certainly take care of that much deflection in a radial sense. The bushing material is pretty elastic and is made to have the sleeve trapped between the x-member flanges.
A friend had his L/F snap off as he set up for that nasty off camber right into Thunder Valley at Mid Ohio. He went straight off and the gravel trap stopped him about 3-4' from the wall.
10-03-2006, 08:41 AM
#53
Mr. Excitement
Rennlist Member
Rennlist Member
The tire hitting something will load up the arm the same as well. Dropping into a pothole or comming back on track over a abrupt sholder. this is also often done with the brakes on thus increasing the spike loads.
Note the oyster shell pattern to most of the metal failure points and that there are 2 distinct shelves or ridges to this pattern. It looks like there was an event that overloaded this area and caused a set of cracks to start. Cracks, not a full failure. This can be noted in the rust that is only present at the starting points in the tube end weld areas. Had this been spotted and corrected the parts might have been easily repaired. Once started the cracks acted as stress risers and concentrators. Once there is enough energy or weakness to permit metal matrix damaging movment the conection is doomed. The next phase of the failure continues with overflex failure that causes the metal matrix at the high stress points to weaken via internaly generated heat and matrix breakdown. The stressed metal continues to flex and loose strength at the stress points until a second event of ever smaller requirement over time occurs and the parts, well, part. This is the second shelf of the oystering in some areas and the rough pulled apart areas in others. The tube sides exhibit classic primary overload / secondary stress riser failure. I also think Timman's insightfull note on bushing binding might have had a part in this. An overload event along with a binding bushing to inject flex into the freshly made stress risers.
As was said the pivot tube can be reinforced with added gussets. I like to drill / weld through holes in full contact overlayments like this. Drill and weld through evenly spaced holes to spread the load around to more than the edges of each part. The edges are more likely to have stress risers that aid in crack formation. The A arm had the body well reinforced to reduce flex but the pivot tube was stock and could have been easily reinforced at the time the rest was done. Drill/weld through the existing metal that the tube was welded to and add small gusset and the transition point. 20 min, 2 oz added metal. As it was built the tube welds that were at the edges did much of the work in front to back loading as the curved metal of the A arm stampings would flex more than the sides as the metals flex. Under high or overload the bushing tube would distort as well concentrating the load at the leading ends of the tube as well. The seam weld to the metal at the ends of the A arm metal that curved around the tube was compromised in value as it was at the end of a curve. Adding a round ring welded to the ends of the tube would increase the reistance to cracking at the tube ends as well.
Note the oyster shell pattern to most of the metal failure points and that there are 2 distinct shelves or ridges to this pattern. It looks like there was an event that overloaded this area and caused a set of cracks to start. Cracks, not a full failure. This can be noted in the rust that is only present at the starting points in the tube end weld areas. Had this been spotted and corrected the parts might have been easily repaired. Once started the cracks acted as stress risers and concentrators. Once there is enough energy or weakness to permit metal matrix damaging movment the conection is doomed. The next phase of the failure continues with overflex failure that causes the metal matrix at the high stress points to weaken via internaly generated heat and matrix breakdown. The stressed metal continues to flex and loose strength at the stress points until a second event of ever smaller requirement over time occurs and the parts, well, part. This is the second shelf of the oystering in some areas and the rough pulled apart areas in others. The tube sides exhibit classic primary overload / secondary stress riser failure. I also think Timman's insightfull note on bushing binding might have had a part in this. An overload event along with a binding bushing to inject flex into the freshly made stress risers.
As was said the pivot tube can be reinforced with added gussets. I like to drill / weld through holes in full contact overlayments like this. Drill and weld through evenly spaced holes to spread the load around to more than the edges of each part. The edges are more likely to have stress risers that aid in crack formation. The A arm had the body well reinforced to reduce flex but the pivot tube was stock and could have been easily reinforced at the time the rest was done. Drill/weld through the existing metal that the tube was welded to and add small gusset and the transition point. 20 min, 2 oz added metal. As it was built the tube welds that were at the edges did much of the work in front to back loading as the curved metal of the A arm stampings would flex more than the sides as the metals flex. Under high or overload the bushing tube would distort as well concentrating the load at the leading ends of the tube as well. The seam weld to the metal at the ends of the A arm metal that curved around the tube was compromised in value as it was at the end of a curve. Adding a round ring welded to the ends of the tube would increase the reistance to cracking at the tube ends as well.
10-03-2006, 01:11 PM
#54
Addict
Rennlist Member
Rennlist Member
Join Date: Jun 2001
Location: Claremont,NH
Posts: 586
Likes: 0
Received 0 Likes
on
0 Posts
Is this the only end that breaks that you guys know of..
I once thought I saw that someone was building up the sway bar area..
or was this done so you can use the later drop links?
I once thought I saw that someone was building up the sway bar area..
or was this done so you can use the later drop links?
10-03-2006, 02:02 PM
#56
Addict
Rennlist Member
Rennlist Member
Hey;
The sway hole will break over time. I have also seen the arms bend at the bar mount, or fatigue and crack through the edges of the arm pieces over a long time from flexure against a stiff spring/strut. These are all exacerbated if the car has been towed using long hooks that tend to damage the profile of the stamped arm halves.
Unreinforced arms are maintenance items.
The sway hole will break over time. I have also seen the arms bend at the bar mount, or fatigue and crack through the edges of the arm pieces over a long time from flexure against a stiff spring/strut. These are all exacerbated if the car has been towed using long hooks that tend to damage the profile of the stamped arm halves.
Unreinforced arms are maintenance items.
10-03-2006, 02:16 PM
#58
Drifting
How about the aluminum arms? I know a lot of people recommend changing to the steel arms for racing but is that because the Al arms are more expensive and the ball joints are more difficult to service or replace, or do the aluminum arms experience similar catastrophic failures? I have Al arms on my car and I have inspected them and found no cracks. Is it just the belief that the steel arms are more ductile and will give you more warning before they fail that drives everyone to recommend them, or are aluminum arms failing at a higher rate than the steel arms?
10-03-2006, 02:21 PM
#59
My understanding is that the aluminum arms are more likely to experience a catastsrohpic failure if the car has been lowered.
Usually, the failure with the steel arms are gradually. I have heard that the failures with the aluminum arms are more likely to be sudden. My understanding is that a lowered car causes too much stress to be put into the aluminum arm at, IIRC, the ball joint.
Usually, the failure with the steel arms are gradually. I have heard that the failures with the aluminum arms are more likely to be sudden. My understanding is that a lowered car causes too much stress to be put into the aluminum arm at, IIRC, the ball joint.
10-03-2006, 02:34 PM
#60
Race Director
The ball joint is the main weak spot on the aluminum arm.
TD is right about catastophic failure. I believe he felt the car being "off" before the arm actually broke. Aluminum will tend to feel fine unit it breaks.
TD was I wrong on this?
TD is right about catastophic failure. I believe he felt the car being "off" before the arm actually broke. Aluminum will tend to feel fine unit it breaks.
TD was I wrong on this?