Just curious, how did this IFS rig do?:D

I was watching that rig working on the top part of the course. He timed out on an obstacle, and broke the rear pumpkin getting pulled out. I think his day was done after that.
Phil

Originally posted by OOP'S
Just curious, how did this IFS rig do?:D

it offered him the oppertunity to do alot of spectating. :D

Originally posted by camo


it offered him the oppertunity to do alot of spectating. :D

Bawahahahahaha :laughing:

Anyone know which IFS rig blew apart the aluminum Tera housing?

it was shannon's rig in the above pic

Originally posted by camo
it was shannon's rig in the above pic

Thanks dude..........oh by the way nice 12 min. axle swap on Johns buggy. :D

team work , air tools and to much practice = quick axle change :D

Originally posted by camo
team work , air tools and to much practice = quick axle change :D

With heavy emphasis on the "too much practice" :D

On friday he also blew apart one of his spiffy halfshafts. sheared it off right at the weld. I was damn surprised, that thing looked freakin beefy.
Dallas

Randy didn't fair any better on some but truly was pick your line as some he crawled over where others used whatever....and Donovan(sp) HP won't get you far in this sport, but hey he put on one hell of a show!

Anyone know which IFS rig blew apart the aluminum Tera housing?
We were wondering about that housing being aluminum or not. Is it a direct copy of a factory housing, just a material change? Didnt look like it had much webbing on the outside for strength.
Phil
P.S. WOOHOO finally got my star!! Thanks Lance!!

after they pushed him ovver from his lay over, he timed out

then, his buddy was winchin him sideways from the roll bar, then the TERA aluimum housing split. thats all i know, my guess was that was his last run:eek:

That rig seemed so promising, was it the aluminum standad Tera 60 or the new CRD?

it was based off a hp tera 60.

And let's not forget mentioning Pat Gremillion's new rig. His is IFS/IRS like Randy's, but a little different. He broke a rear steering stop on A5 the first day. Not a major breakage, but enough to put him out for the day. His scorecard wasn't so great the second day but it held together. Randy's broke twice that i saw. He actually broke a halfshaft after the exit gate on B7. He got through the obstacle pretty clean, then got hung up, stuck, and popped a shaft!!

Ohh well. The indy rigs fared poorly in JV...

Originally posted by StinkBug
On friday he also blew apart one of his spiffy halfshafts. sheared it off right at the weld. I was damn surprised, that thing looked freakin beefy.
Dallas

He had a *welded* halfshaft? Is it any surprise it broke?

--Dan

Originally posted by houlster


He had a *welded* halfshaft? Is it any surprise it broke?

--Dan

its not a halfshaft in the traditional sense, it was more like a really short drive shaft with long slip splines and 1ton yokes on each end. I have a pic around here somewhere i'll post.

Dallas

http://66.12.181.251/2002rcaafinals/pict1049.jpg

A little problem with the rear axle...
http://66.12.181.251/2002rcaafinals/pict1078.jpg

Seemed to get over the rocks well...
http://66.12.181.251/2002rcaafinals/0997.jpg

i like how they did the skidplating on that rig

here are my close ups of the halfshafts



unbroken shaft (http://community.webshots.com/photo/50242072/50908851hMZoqE)

broken shaft (http://community.webshots.com/photo/50242072/50908892OEHsmj)

Dallas

It would be interesting to find out why it broke. I can't really tell. There isn't much travel there though. Wonder if he has any issues with the splines binding. We ran some similar ones on a old Baja car and had nothing but problems with them binding. Since have gone to CV joints instead. And that aluminum diff break is sick.

[A little problem with the rear axle...
http://66.12.181.251/2002rcaafinals/pict1078.jpg
]

AWSOME BREAKAGE!!!!!!!!!!!!!!!!!!!! I LOVE IT....

Yep Lance told me about that break last night................All Aluminum Housing...........:eek: Safe to say there day ended right there.

Originally posted by Weasel
It would be interesting to find out why it broke.

the weld broke I think lance said. i believe that's a hi-angle driveline sticker on it in the first picture.

- jack

I love it. all that helical gear side load.
aluminum has 1/3 the elasticity of steel.
I'll bet that crack starts on the inside of the bearing caps.
where is Tera Mfg? I'll send them my freshman materials text book

The weld broke?? Seems hard to belive since I know high angle makes good stuff but to break a weld. Wouldn't that be a pretty crappy weld?

Jess makes good stuff no doubt.

but there is alot more stress involved when asked to be a driveshaft versus being a drive axle I'd think. say he had 4.88 gears multiply whatever torque is coming out of the tranny by 5
and the numbers get big quick ....

Jess said a few weeks ago that he'd only ever had one driveshaft fail due to a weld.

- jack

Jess did not do the weld. And the weld did not break, the shaft sheared right next to the weld. Shannon told me that they should have heat-treated the shaft after welding it, but they didn't have time before the comp.

Just to set the record straight.

Thats spectacular breakage :eek:

Originally posted by Sundowner
I love it. all that helical gear side load.
aluminum has 1/3 the elasticity of steel.
I'll bet that crack starts on the inside of the bearing caps.
where is Tera Mfg? I'll send them my freshman materials text book

That's what I was thinking too. Looks like they made it EXACTLY the same as the iron version. No consideration for different material at all.

Broken aluminum diff, I never saw that coming:rolleyes: It was a great idea to cut out extra weight but it just dont seem that smart to me

Originally posted by Sundowner

aluminum has 1/3 the elasticity of steel.


Its also 1/3 of the weight, looks like the gamble didn't pay off.....

Originally posted by elf_cruiser
Jess did not do the weld. And the weld did not break, the shaft sheared right next to the weld. Shannon told me that they should have heat-treated the shaft after welding it, but they didn't have time before the comp.

Just to set the record straight.

Thanks for clearing that up. I was finding hard to belive Jess had anything to do with it.

Would welding a shaft that has already been heat treated weaken even more? I don't know if heat treating it again would help.

Would welding a shaft that has already been heat treated weaken even more? I don't know if heat treating it again would help.

I am far from being a metallurgist, but my basic understanding is that welding on heat treated steel creates a weak spot around the weld. A post heat-treat will even out the stress from the weld, it kinda realigns the molecules, and restores the original strength...

Hopefully someone else can explain that a little better!

the housing didnt break from wheelin it hard, it broke when they were winching him out, there was a rock at the bottom of the diff, and they were pulling down on it, and it broke.

like taking a stick and breaking it over your knee:rolleyes:

I saw the front axle break and anything would have broke it that situation. It was wedged in under a ledge.

Originally posted by elf_cruiser


I am far from being a metallurgist, but my basic understanding is that welding on heat treated steel creates a weak spot around the weld. A post heat-treat will even out the stress from the weld, it kinda realigns the molecules, and restores the original strength...

Hopefully someone else can explain that a little better!

You are on the right track elf dude. Welding on heat treated stuff or anything for that matter will cause a difference in crystal structure in the metal. Because the HAZ will heat up and cool fast causing hardness which in turn causes brittleness. The degree to which this occurs depends upon the base material and if it is heat treated (rockwell #). Annealing or post heat treating will help this by allowing the crystal structure to reallign the same throughout the material. Got it...? Oh yea and elf, you drink like a girl.:flipoff2:

Originally posted by morpheus
Jess makes good stuff no doubt.

but there is alot more stress involved when asked to be a driveshaft versus being a drive axle I'd think.

I think you said that backwards of what you meant.

--Dan

It was a great idea to cut out extra weight but it just dont seem that smart to me

I don't see the point in trying to save the weight at the axles.
once you get into 35" flotation tires on an Al axle, it's like supersizing the double quarter pounder meal and getting a diet coke to go with it.

Originally posted by Sundowner


I don't see the point in trying to save the weight at the axles.
once you get into 35" flotation tires on an Al axle, it's like supersizing the double quarter pounder meal and getting a diet coke to go with it.

Haha....that was funny!

Oh yea and elf, you drink like a girl.


WTF!!??? At least I was there until the end of the party!!

You passed out like a little girl - BITCH!!!

Originally posted by elf_cruiser



WTF!!??? At least I was there until the end of the party!!

You passed out like a little girl - BITCH!!!

OK dude, I wanzed out on fri. night (I had a late night thurs.). Where the hell were you sat. ????? What the hell happened to this thread? It went from shit talkin IFS to shit talkin.... I guess I started it :flipoff2:

I begged Ranch (from Tera) at Sema last November for a Alum. Tera 60 and he basicly said no. He didn't want anyone to have one that would be used for competion because if it broke, everyone would see. I was pissed that I could talk him out of any.

Once again Ranch was correct....everyone say.......just not me....

see ya,

tony k

it kinda realigns the molecules

Pig already said it, but just a trivial FYI, usefull if, say, MR4WD fancies himself a metalurgist amongst his many other talents: there is no such thing as a "molecule" of steel....the expression you are looking for is the "crystal lattice structure":flipoff2:

Originally posted by BillaVista


Pig already said it, but just a trivial FYI, usefull if, say, MR4WD fancies himself a metalurgist amongst his many other talents: there is no such thing as a "molecule" of steel....the expression you are looking for is the "crystal lattice structure":flipoff2:

Disclaimer---- I am not a Metalurgist either!!!!!

Well if you want to get technical the word you are looking for is martensite (sp?). The base material adjacent to the weld puddle is rapidly quenched at the heat effected zone causing the formation of martensite. This phase is very hard and brittle. It also has low fracture toughness. What that means is if there is a small surface flaw and a shock load the part will have rapid crack growth typically resulting in full section failure, especially under torsion. Removing surface flaws by grinding the surface will significantly decrease this effect.

However there is another factor that needs to be acccounted for in welding. The tensile residual stress left when the weld puddle cools. That typically results in stresses at or near the material yield strength. Anealling will reduce the residual stress state which in effect increases the shock load capability of the weld joint.

Disclaimer---- I am not a Metalurgist either!!!!!

WTF!!! I understand BillaVista's statement about crystal lattice structure, but i didn't understand any of what you just laid down. I think you need to speak ENGLISH from now on...

:confused: I used as many small words as I could!

Sorry, sometimes only a part of what's in my head comes out on the key board. I'll try to explain it a little better. The problem is that I work with welded structure fatigue on a daily basis. I get used to the terms and assume other people are familiar with them also.

Here goes, if parts of this don't make sense, or all of it, just ask for clarification.

There are many phases or lattice structures that steel can form as it's cooling. The amount of carbon, the starting temperature, and the rate the steel is cooled control what phase the steel ends up. When you look at the part you can't see the difference but under a microscope it's very different.

The most common phases are pearlite, austenite and martensite, although there are other phases. Martensite is formed when you take a steel with adequate carbon content from molten and quench it very quickly.

This happens when you weld. The material around the weld is cold and absorbs the heat from the molten weld puddle very quickly (quench). The weld puddle solidifies from the outside in. The heat affected zone is the portion of the weld that cools the fastest (farthest outside) and therefore typically forms martensite. Preheating and post heating control the quench rate of the weld by slowing down the time it takes for the weld to solidifyat the weld toe giving a less brittle transition between the weld and the base material.

As for residual tensile stresses. These are the result of the weld puddle cooling. The weld puddle takes up the most volume when it's hot. As it cools it shrinks and tries to pull the material around it in tighter. The material around it is already cooled and solie and it tries to pull back from the weld are which forms these residual stresses. Because the weld shrinks so much it can cause very high stresses and for design of welded structures we assume the residual stresses at the weld toe are at the material tensile yield stress. In simple terms tensile stress (positive) can be thought of as "pulling" on the material. Yield stress is the point when you're pullin on a part and it won't go back to it's original state (permanent deformation).

Forging on the other hand can create surface compressive stresses. These compressive stresses are caused by forcing the material to flow over the dies and it also helps align the grain structure and reshape it locally. Compressive stresses (negative) can be thought of as pushing on the material.

Why do residual stresses have an effect on how much load a part can see before it fails? Lets assume the external load causes a tensile stress in the part. For the first example of a welded piece the residual stress adds to the external load and the part yields further along the stress strain curve. If enough external load is applied it will actually fail.

For the other part that has a compressive residual stress, the external load first has to overcome the compressive stress, then the part starts to see the tensile stress applied to it. This part will take a MUCH higher external load to cause failure.

Another question is how to eliminate or reduce residual stresses.

Reheating the parts in an oven for several hours to several days depending on the size of the part will reduce the residual stress. The temperate and rate of temperature change both need to be controlled properly to accomplish this.

What I described above holds true for most situations but when the external load is applied really fast (shock load), like when a tire hops or slips and catches these have less of an effect.
Shock loads are really just very fast accellerations or decelerations. But the equations for mostly static loads don't necessarily apply directly. There is a whole field of study called fracture mechanics that gets into the study of crack growth.

What it boils down to is very small surface flaws, stress concentrators, low fracture toughness, and high shock loads will break parts very quickly as we've all learned the hard way.

Welding creates surface flaws. I won't get into all of the issues but there is in essence a crack at the toe of the weld. The only way to eliminate that crack is to grind the surface smooth. You need to take out at least .5mm but preferably 1-1.5mm at the weld toe. Take away that crack and the part can withstand much higher shock loads. Another positive is grinding also relieves some of the residual tensile stress in the weld.

Finally :flipoff2: if you can't understand plain english. :p

you talk like one a dem stoopid engineer types.
God, I hate engineers..;)

That bieng said,

the last American Institiute of Steel Construction Seminar I went to indicatated that if you grind flush the toe of the weld and peen it, you can expect 50-70% better performance from the weld in repetitive loading situation. the term they used for the impurities in the toe of the weld was "microslag"
if you don't grind it out, it's kinda like when you take a bite out of the side of a strip of packing tape to make it break off.

WOW that was some great tech. I wish i had read that when i was in HS still. i coulda printed it out and gotten an A+ on any number of papers in quite a few classes :flipoff2:

Dallas

We use the term slag intrusion but it's the same theory.

We use a factor of 2 or 2.2 for machine type grinding, not by hand. A hand held grinder with a soft backer and 100 grit paper would be a 50-75% improvement for cyclic loading. Also grinding the weld toe flush will not eliminate the slag intrusion. It has to be below the surface by at least .5mm but again preferably 1 - 1.5mm.

There is no rule of thumb for shock loads. The life improvement could be 1% or 50 times. Just depends on the factors I mentioned earlier.

Step on soap box here.

Yea, I am an engineer, and I've run across many that had no business in their job. A lot of the guys I work with don't do any of their own car repair let alone build anything.

I had a situation a few years ago where a guy in another group was designing a part for a vehicle I was responsible for. He designed a bracket that couldn't be removed without first taking off 3 50lb filters. All he had to do was move the bolts over 2" either way and no problem. I finally convinced him to correct it when I told him he would have to go out in the field the first time we had a failure and take the parts off himself.

I think we would all be better off if the designers were forced to repair or assemble the products they design for some period of time.

Another pet peeve is when engineers design stuff without getting input from the people that madhine, assemble, repair and rebuild the stuff. Those are the people that have the first hand knowledge that is critical to making it easier to machine, assemble, repair, or replace and in the end makes it cheaper for the customer.

I better get off my soap box.

you're preaching to the choir here :-)
I love teliing other engineers that YES, that's a beautiful design, but how were planning on getting a wrench in there to tighten the bolts. or my personal favorite: that's a nice beam, too bad thay don't make a crane that can lift it.

It1yj: Very good tech. It reminded me of my Materials Eng studies of a few years back. damn I loved that stuff, too bad everything else sucked. All this talk about martinsite got me to recall the old iron carbon phase diagram. This might help some people that are actuallty reading this, that want to learn.
http://www.sv.vt.edu/classes/MSE2094_NoteBook/96ClassProj/examples/FeC.gif

I found a neat picture of martinesite. Look how close and smashed together the crystal structure is.
http://info.lu.farmingdale.edu/depts/met/met205/martensite.jpg

Compare that to austinite which occurs lower on the Iron Carbon phase diagram.
http://info.lu.farmingdale.edu/depts/met/met205/austenite.JPG

my brain hurts now.:(

If someone could double check this. I am pretty sure that this is where the Marensite occurs but could not find any documentation.

PIG - "right click, save as..."

LT1yj - thanx a lot dude, that was fawkin cool!

ah, those diagrams bring back some memories of last year. I also got a cooling chart around here somewhere, has to do with how fast you quench the steel and what it forms.

Ahhhhhhh
This is good. Kind of turned away from the original post but very good material and weld tech! A guy could spend some time with the diagram looking at what to do when welding up a cast iron diff housing also.....

yeah thanks for the time to post this stuff guys, Its so nice to have some tech instead of that wining bithing stuff that has been here lately ..hehe

Originally posted by Moab Austin
yeah thanks for the time to post this stuff guys, Its so nice to have some tech instead of that wining bithing stuff that has been here lately ..hehe

Yeah, I agree completely. Thanks!

Originally posted by PIG
I found a neat picture of martinesite. Look how close and smashed together the crystal structure is.
http://info.lu.farmingdale.edu/depts/met/met205/martensite.jpg



Actually I belive the lines that you see are line dislocation or places where extra half planes of atoms have been shoved into the crystal latice. Dislocation are normal good becuase the strain the lattice and makes in stronger. But as you apply a tensile load it cause the dislocation to travel in a direction(Burgers vector for all the techies). As more load is applied these dislocation will start to collide and get tangled up in each other as in the picture and then you start to get plastic deformation of the metal.

It's been more than a few years since I studied a steel and iron phase change chart. This is a good chart (thanks) as it reflects the eutectic threshold for low and high carbon steel (the temperature and carbon content to anneal pearlite and ferrite or cementite).

The 1333 dF threshold, where Austentite begins to form is when the cherry red steel starts to grow and "dance" (BCC to FCC phase change), and the 1414 dF threshold is when the cherry red steel loses it's magnetic property (IIRC). These were valuable indicators to primitive blacksmiths when altering steel for various usages, and can be used by everyone to improve steel products without the technical instrumentation. You do not need to know the technical terms, just that you need to heat the steel to the point is grows and "dances fire" (or loses it's magnetic property) before quenching or annealing slowly to gain the hardness desired.

Two other rough indicators for clean steel (IIRC) are a blue color for temperatures over 450 dF. and a straw color for temperatures over 1200 dF. Blacksmiths used these indicators to allow them to control the heat during post fabrication annealing (thermometers were hard to find in pioneer times). Typical usage called for Austintite quenching, followed by dressing the part clean and post fab heat-soaking at a straw color for hardened steel and a blue color for mallable steel (before additional quenching). Heat soak monitoring, watching the steel color and the fire, for hours, was apprentice duty. The soak times were determined by skilled Smith testing and experience.

New steel shipments in pioneer times would require a long test rod to be heated into Austintite, quenched and dressed clean, and then heated on one end for annealing under a timed condition. The color change indicated, and the resulting hardness along the piece, would provide the experienced Smith an indicator of the quality of the shipment (the carbon content) and how to work the steel. The result would provide a crude version of an instrumented indicator of the steel quality, simular to the result of an ASTM Jominy (sP?) Bar test.

Now, since the failed diff housing was aluminum where is the polyphase change chart for 6061-T6, or the material spec for the housing? What temperature for heat soaking would have restored the lattice structure for the doped alloy in the aluminum (the grain structure that could have been altered during welding or machine induced heat stress).

Funny, the last time I saw a non-ferris diff fail was a magnesium 8 3/4 Chrysler, one that spun the pinion (bearings and all) out of the dropout casting (the reason I looked past the first page of this thread). This failure looks completely different, regardless of how the failure occured, and I am interested in knowing what the internals looked like (did the failure enter the bearing caps)?

Thanks, again, for the chart (and stimulating long dormant brain cells). If someone can confirm the blacksmith indicators (I am going from pure lost memory) it might help the "getto-fab" steel products improve.

Happy Trails!

I like this thread, the idiots have shut up and the sensible and intelligent people can share some info to learn something usefull. I'd love to chime in with some info but got some big projects to finish for SEMA.

Originally posted by Ed A. Stevens

Now, since the failed diff housing was aluminum where is the polyphase change chart for 6061-T6, or the material spec for the housing?

What kind of material specs did you want?

Here's a cool link I found. Lots of info about aluminum.

www.engr.ku.edu/~rhale/ae510/aluminum.pdf

Need Adobe Acrobat to view it though.

What was the failed housing made out of, the aluminum class and treatment? I apologize if was given already, I missed reading it in the thread.

Happy Trails!

Originally posted by lt1yj

Welding creates surface flaws. I won't get into all of the issues but there is in essence a crack at the toe of the weld. The only way to eliminate that crack is to grind the surface smooth. You need to take out at least .5mm but preferably 1-1.5mm at the weld toe. Take away that crack and the part can withstand much higher shock loads. Another positive is grinding also relieves some of the residual tensile stress in the weld.

Finally :flipoff2: if you can't understand plain english. :p

I'm one of the idiots.

This (http://www.adtdl.army.mil/cgi-bin/atdl.dll/tc/9-237/Ch6.htm#s5) just made a lot more sense. It's the Army welding manual, and it might be useful in this thread. If you search 'residual stress', it will take you to exactly what they're talking about.

Where could I acquire a thermometer to follow these techniques? .

One point I forgot to mention is a properly sized/designed weld joint will fail at the weld toe. Failures through the face or throat of the weld usually indicate undersized or incorrectly applied weld.

A little more praise for this thread. Truely "College Level".

I am one of those machinist that has been on the recieving end of those pretty, but not very feasible designs. I had enough and am working on my BS degree in ME at 31 with 3 kids to raise. One day it'll pay off for all of us!

Goat1, SEMA?!?! Cool, dream job of mine after school would be anything that had a project for SEMA or SHOT.

Time to dream about i hat, j hat, and k hat while I sleep....

Most informative thread I have read in a while! Thanks guys! :beer: :beer:

One point I forgot to mention is a properly sized/designed weld joint will fail at the weld toe. Failures through the face or throat of the weld usually indicate undersized or incorrectly applied weld.

that depends on how the weld is loaded.
welds are best for pure shear and the rod metal is usually 20-40ksi stronger than the base metal. you start loading them in tension, build circumferential welds, and then you get into Blodgett's equasions and you can lose the base metal a lot faster. I've seen it happen a few times in high fatigue situations. The newest designs we do are starting to use matched or under matched welds. this has been true of the hybrid sections that have elements of different yield strengths welded toghether

Originally posted by Sundowner


that depends on how the weld is loaded.
welds are best for pure shear and the rod metal is usually 20-40ksi stronger than the base metal. you start loading them in tension, build circumferential welds, and then you get into Blodgett's equasions and you can lose the base metal a lot faster. I've seen it happen a few times in high fatigue situations. The newest designs we do are starting to use matched or under matched welds. this has been true of the hybrid sections that have elements of different yield strengths welded toghether


You're right, I oversimplified. When loading is applied in a pure axial direction to a weld the weld acts very similarly to parent material.

However, it's very rare for a weld throat to fail before the weld toe in high cycle fatigue. The slope of the SN curve for a weld is -3 and the parent material can be -6 to -11. In addition the welds typically occur in transition regions that create a stress concentration factor at the weld toe. In most fabricated structures the loading is very complex and almost never occurs in a pure axial direction.

Low cycle fatigue has a higher probability of weld throat failure because its operating close to the material yield strength. Any inclusion or flaw in the weld will grow rapidly and typically the parent material has smaller inclusions.

Getting into matched or undermatched welds would increase the weld throat failure rate. Especially in low cycle fatigue.

One good example of section failure (not throat) would be plug welds (rosette) on the axle tube to center section of an axle assembly. These welds see primarily shear stresse and the weld toe is typically not an issue.

Can you think of another example that would apply to offroad vehicles?

I think or respective disciplines are defining heigh-end fatigue differently. For me, I'm looking at 2-5 impacts/sec over a 50-75 year life span. you seem more biased towards machine design situations.

Getting into matched or undermatched welds would increase the weld throat failure rate. Especially in low cycle fatigue.

Undermatched welds have been useful (for us) with preventing HAZ hardening and hydrogen cracking in the high strength, low alloy metals.

One good example of section failure (not throat) would be plug welds (rosette) on the axle tube to center section of an axle assembly. These welds see primarily shear stresse and the weld toe is typically not an issue.

Iv'e always felt that axle plug welds failed through the shear plae in the section due to the different elasticities of the steel tube and the cast housing. The tube strain hardens the plug weld at the interface plane, and the cracks propogate through the section.

As far as other offroad applications, we should mention that mixing bolts and welds in a connection is not a good idea. theweld remains static for its loading and the bolts cannont mobilize to develop shear bearing on the connected elements or slip critical restraint due to friction

Originally posted by Sundowner

Iv'e always felt that axle plug welds failed through the shear plae in the section due to the different elasticities of the steel tube and the cast housing. The tube strain hardens the plug weld at the interface plane, and the cracks propogate through the section.



The idea behind the pressed tube with the rosette welds is to eliminate the weld in a transition region. The ductile center section carries the bending load and the combination of press fit and rosette welds carry the torsional loads. The high stress region at the end of the casting on the tube is parent material, instead of a weld. The parent material will live orders of magnitude higher than a weld in the same place.

I don't think the material elestacities have any impact on the weld failure. The HAZ will be very hard since it draws carbon from the cast housing, but in addition the residual stress is very high at the weld, again I assume it's at the yield stress. High torsional loads have to exceed the press fit torque before the weld sees any higher stress. Much more of an issue with lower gears and shock loads.

There is another factor at the plug weld interface. The stress concentration factor is huge at the shear plane between the tube and the housing because of the close fit between the parts.

We categorize high cycle fatigue as above 2 million cycles. Many people claim there is a knee in the S-N curve around there but in reality most of the data doesn't go too far past 10 million cycles. From working on thin sheet metal structures that have natural frequencies in the 50 Hz range and having cracks grow in a few thousand hours I firmly believe the curve just takes on a different slope but I have no emperical basis to justify it.

Sundowner: I take it you're a civil engineer?

I'm a mechanical engineer. I have my own company (actually owned by my wife, "Yes dear") that specializes in design and analysis of structures with a slight bias towards fatigue of welded plate structures. Primarily in the off highway and earthmoving equipment market. Thus my bias towards machine design.

Thanks for the discussion!

great tech but you guys are making me scared to go weld. :D

pig
thanks for the chart. that clears everything up. LOL

Thanks guys for the excellant and quite interesting reading.

NoRM

Sundowner: I take it you're a civil engineer?

I'm a mechanical engineer. I have my own company (actually owned by my wife, "Yes dear") that specializes in design and analysis of structures with a slight bias towards fatigue of welded plate structures. Primarily in the off highway and earthmoving equipment market. Thus my bias towards machine design.


'fraid I am a Civil, well Structural, whatever.
I do bridges and I take a special interst in the forensics of the broken ones.



Thanks for the discussion!

best technical discussion I've had in a while, including here at work.

Camo, I get paid by the hour;)

Originally posted by Sundowner



best technical discussion I've had in a while, including here at work.



Me too! I work by myself most of the time and it's nice to stretch the ol brain once in a while. Have you ever built a "finite element" mesh? Talk about mind numbing.

Camo:

It's unlikely most of us will get in the high cycle fatigue realm where most of this is applicable.

Most of us are more concerned with designing below the yield stress so our parts don't bend and getting a good weld so they don't fall off when we're on the trail.

Have you ever built a "finite element" mesh?

that's what I'm doing right now!
I'm taking sanity breaks with the rest of the freaks here:D

I just wrapped one up and have a little free time before I start the next. The one I finished was 300,000 nodes and approximately a million degrees of freedom, all hand mesh with weld geometry explicitly modeled.

The one I did right before that was 120,000 nodes and 400,000 degrees of freedom. This was a model of the entire strut assembly using 7 nonlinear contact pairs and includes nonlinear springs for the individual rolling elements in the tapered roller bearings. It only took 7 hours to solve in Abaqus with 4 load cases.

What program do you use for pre/post, solutions and fatigue life?

We use SDRC I-Deas, Abaqus and FE-Safe. We throw ProE in for solid modeling.

our software is nowhere near what you're using.
most of it is in-house based, and most of our models we run on STRUDL. I can't imagine how ong that model would take to run. the last BIG one I made wasn't even a tenth that size and it took 5 hours to run on the VAX(yes, a VAX)
We don't have a lot of degrees of freedom with the bridge models, at most, I might release a few dozen for rotation and translation in some seismic loading. most of my soil springs are non linear depending hon how competant the geotech is.

ProE? I havn't used that since School. I take it you're using a Sun system for those models?

how can you model weld geometry? you must have strict enforcement of AWS standards to be able to get it accurate.

Originally posted by lt1yj


Camo:

It's unlikely most of us will get in the high cycle fatigue realm where most of this is applicable.

Most of us are more concerned with designing below the yield stress so our parts don't bend and getting a good weld so they don't fall off when we're on the trail.

so in laymens term................is it safe for me to go weld something on my rig? or am i jepordising the mocular integrity of it its martin thingy my bob ? when i weld i just turn a few knobs and pull the trigger. you guys are scaring the crap outta me.

so in laymens term................is it safe for me to go weld something on my rig? or am i jepordising the mocular integrity of it its martin thingy my bob ? when i weld i just turn a few knobs and pull the trigger. you guys are scaring the crap outta me.

camo, most of the stuff we deal with on a daily basis is a couple of orders of magnitude greater than we will ever see on the trail.
here at work, I consider anything under 100,000lbs to be a small load.
you put a rubberband around a newspaper and don't think twice about it, you put a rubberband around a couple of steel bars and you start to wince a little. that's when the calculators come out.:D

ok all you garage fabricators it is safe to go back to constructing your rigs. just remember to use the welder and not the elmers glue bottle. :D


the rest of you smart guys carry on with your most enlighting discussion and us lay people will try and get up to speed.

Originally posted by Sundowner
our software is nowhere near what you're using.
most of it is in-house based, and most of our models we run on STRUDL. I can't imagine how ong that model would take to run. the last BIG one I made wasn't even a tenth that size and it took 5 hours to run on the VAX(yes, a VAX)
We don't have a lot of degrees of freedom with the bridge models, at most, I might release a few dozen for rotation and translation in some seismic loading. most of my soil springs are non linear depending hon how competant the geotech is.

ProE? I havn't used that since School. I take it you're using a Sun system for those models?

how can you model weld geometry? you must have strict enforcement of AWS standards to be able to get it accurate.

We model "nominal" geometry. The weld size specified on the print. Typically the weld volume is greater than what's on the print but the weld volume increase doesn't change the stress state very much. In addition we only look at the stress 1 plate thickness away from the toe of the weld. That way if the weld geometry isn't exactly right it has no effect on the life calculation. We are using BWI S-N curves and weld classifications.

Believe it or not, I'm running a Dell 620 and 530 workstations running Win2K. I have a 1.7Ghz processor, 4Gb of Ram, and 4 18Gb scsi disks using a RAID controller to access all 4 disks simultaneously. This setup is faster than my customers HP Unix workstations with comparable hardware (which are probably 1-2 years old).

As for DOF's in the model. Each node has a total of 6 dof's to start with. Depending on the type of element some of the dof's drop out of the equation. Linear solid elements have 3 dof's, most plate elements have 5 dof's, beam elements have 6 dofs per node. The whole component will only have 6 free body dof's. 3 translation, and 3 rotation. Different points on the structure have to be constrained to eliminate those 6 dof's. The dof's I was talking about are the 3 dofs per node for a solid element times the number of nodes in the model. There are a few more because the mesh I ran last time was a hybrid plate/brick mesh.


Originally posted by camo

so in laymens term................is it safe for me to go weld something on my rig? or am i jepordising the mocular integrity of it its martin thingy my bob ? when i weld i just turn a few knobs and pull the trigger. you guys are scaring the crap outta me.

In laymens terms: yes it's safe in most cases.

There are some areas I feel it is EXTREMELY unsafe. For instance welding on pitman arms, steering castings or some steering linkages. The reason this is unsafe is these components are under high stress, they can be cast iron, ductile iron, nodular iron, cast steel or forged steel, or plate steel and may require special treatment. They also control the vehicle and anyone without excellent mechanical abilities shouldn't mess with them.

I was with a guy at Attica last year in a TLC. We were playing in the bowl climbing the gravel piles and then went over to the quary. The first rock he climbed over his whole steering arm fell off and he couldn't steer. If that would have happened climbing the piles he easily could have rolled. He had welded this funky lift bracket on to raise his drag link to get it somewhat parallel to the ground.

one last note, then I'll shut up:
I wouldn't advise anyone to weld in engine mounts, either.
it's high stress and high vibration.

gotta hold the motor in place somehow :D

- jack

Most of the discussion is over my head, though it is fun to try to wrap my piddly experience around it. My background with Altair Engineering's (did documentation for them so many moons ago), HyperMesh, Motionview, and OptiStruct (now THAT was some cool code!)... talk about a rusty skillset!

As far as other offroad applications, we should mention that mixing bolts and welds in a connection is not a good idea. theweld remains static for its loading and the bolts cannont mobilize to develop shear bearing on the connected elements or slip critical restraint due to friction.
Sundowner, this is interesting. I've always considered bolts and welds a belt-and-suspender solution. Are you saying that the weld will have to fail first for the fastener to see load... or are you citing the different characteristics of the joints, even unloaded, and suggesting that interaction of these different characteristics could lead to failure?

Yield failure on the weld should be much higher than the fastener (unless you're using Grade 12 bolts :groan:), but I still like the idea of bolt-backup on failure-critical junctions...

Randii

I wouldn't advise anyone to weld in engine mounts, either.
it's high stress and high vibration.
Dagnabbit.. there are welds on both sides of my engine mount rubbers... better run a bead of Elmer's glue down them to reinforce a bit... :eek:

I'm with ya randii ... better get the phase particulate accumulator fired up and go realign the molecules before my motor falls out :D

- jack

I welded mine in.

The stress amplitude from the powertrain is fairly low. Accelerations can create high stress but as long as the mount is fairly wide and gusseted it shouldn't be a problem.

I do check mine for cracks a couple times a year.

wow good thread. How did I miss this one. Anyway since we are talking about what we do I'm an R&D engineer in the medical device industry. I look at welds, and fatigue loading of parts with .003-.020 in cross sections, so the stuff I do is on the opposite end of the spectrum as most of the other engineers on here.

surprised no one go this one yet... Martensite is not on the phase diagram because it is not a stable phase. You got to look at a TTT diagram for that. That is one of my favorite interview questions for metalurgists.

Originally posted by PIG
If someone could double check this. I am pretty sure that this is where the Marensite occurs but could not find any documentation.

If they miss that one I give em one more shot. Here it is. You are building a Race car the rules place most of the tubes in the chassis and specify the wall thickness and diameter and that it be a "magnetic" steel. You do some simple analysis and realize that no yielding will occur regardless of what alloy you use. You want to maximize the stiffness of the chassis so what alloy do you pick. 8620 4130 1020 or A36, and how would you heat treat it to maximize the stiffness?

It is amazing how many degreed metalurgists don't get it.

Originally posted by Gordon
wow good thread. How did I miss this one. Anyway since we are talking about what we do I'm an R&D engineer in the medical device industry. I look at welds, and fatigue loading of parts with .003-.020 in cross sections, so the stuff I do is on the opposite end of the spectrum as most of the other engineers on here.

surprised no one go this one yet... Martensite is not on the phase diagram because it is not a stable phase. You got to look at a TTT diagram for that. That is one of my favorite interview questions for metalurgists.



If they miss that one I give em one more shot. Here it is. You are building a Race car the rules place most of the tubes in the chassis and specify the wall thickness and diameter and that it be a "magnetic" steel. You do some simple analysis and realize that no yielding will occur regardless of what alloy you use. You want to maximize the stiffness of the chassis so what alloy do you pick. 8620 4130 1020 or A36, and how would you heat treat it to maximize the stiffness?

It is amazing how many degreed metalurgists don't get it.

Thanks for clearing that up. I was waiting for someone to respond.

You want to maximize the stiffness of the chassis so what alloy do you pick. 8620 4130 1020 or A36, and how would you heat treat it to maximize the stiffness?

OK, I'll stab at this one - I pick 4130 as the material because it has the same elongation as 1020, 15%, but it has a higher ultimate strength, 100kpsi, so it should be stiffer right??

I dunno jack about heat treating, so maybe the heat treat is the catch on this one?? LIke you can heat treat the 1020 to have an elongation of aonly 10% or something, i dunno?

What do I win?

Ummm... Modulus of Elasticity?

I'll put my answer in the test forum. It will be the only post there under my name so you can look when you are ready to give up. It is a simple question to ask a metalurgist, but to someone not familiar with looking at material properties it may seem like a trick question.

I would go with A36, Air hardening tool steel, right? Use a Rosebud tip heat and let "air harden".

Materials aren't until next year in my defense, if I need a defense. :D