Mild steel or 4130 for a race car space frame ? 1

[tensor47]

I am designing/building a small open wheel hill climb car (bitumen track) with a 1 litre motor cycle engine behind the driver. The driver volume will be a CF tub and the engine volume will be a space frame.
I built one in ’68 from ¾” 1/16” wall mild steel and it is still very successful today with the new owner.

The issues I need to consider are :-

1020 and 4130 have the same E so there is no stiffness advantage when using the same sizes.

If I use 4130 and can successfully normalize the joint area after welding to the same as the rest of the frame the whole yield strength is about 63,300 psi.
If I use 1020 the cold worked yield strength of 55,100 psi remains away from the welds, but could drop to 29,700 psi at the welds for annealed (assuming cold forming strength is destroyed by welding) mild steel.
If this reasoning is correct the frame made of 4130 will survive a light crash better than mild steel with same sizes.


My question relates to the choice of mild steel of 4130 from the space frame.

It is too large to be put in a furnace and would distort if quenched in water (along with any holding jig), and would need tempering even if all the previous was possible.

I would be very happy if I could end up with a minimally distorted, fully normalized 4130 frame.

My realistic choices are Tig welded mild steel or Tig welded 4130 or nickel bronze brazed 4130.

Mild steel option.
Q1
Cold drawn is a lot stronger than annealed but will the Tig welding heat reduce the strength gained from the cold drawing ?

Q2
I am assuming no pre of post heating in Q1, but is this necessary ?

4130 option.
Q3
If I oxy torch heated an as purchased “normalized” 12 length of ¾” dia 1/16” wall tube in the middle 4” to the ideal normalizing temp and air cooled it in still air, would the whole length of tube be normalized as purchased ?
Or would there be HAZ areas near the end of my heating zone ?

Q4
At a typical joint there is a long horizontal tube, mid way along a vertical tube butted to it, a horizontal tube butted to it, and 2 horizontal diagonals plus one vertical diagonal all fitted to it.

Q4a (with 4130 filler rod for consistency throughout)
If I sequentially weld all these joints, which increases the mass per unit surface area, and then attempt to normalize it with an oxy torch, will the whole joint and the rest of the frame be normalized ?
Will there be any HAZ areas ?
Will the generally denser volume be softer ?

Q4b (with ER-70 S2 or ER-80 S2 filler rod for realistic joint ductility)
If I sequentially weld all these joints, which increases the mass per unit surface area, and then attempt to normalize it with an oxy torch, will the whole joint and the rest of the frame be normalized ?
Will there be any HAZ issues ?
Will the generally denser volume be softer ?

Q5
I read this article recently about cracks opening and filling with brazing material.

"Always avoid brazing 4130 steel. The reason to not braze chromemoly is that this steel has a definite grain structure that actually opens up at medium red brazing temperatures. When brazing alloy is melted onto the steel surface, it flows easily into the many small cracks and crevices in the chromemoly steel. Then, as the braze joint cools, the brass will not compress and it forces major cracks to form in the 4130 steel. Often, a brazed 4130 steel part will crack completely in two before your eyes as it cools."
Performance Welding, Richard Finch, c. 1997, pg. 130.

I have seen mild steel tubes crack on cooling after overheating and burning the bronze accidentally.
Is this a similar ?

Cheers

John

 

[Metalguy]

If you can use 4130 and gas weld it. You do NOT need to "normalize" it after welding. You just don't want it to cool fast during welding, so use the gas flame to help maintain some heat in the weld area.

The critical temperature for 4130 during cooldown is right around 1200 deg. F., which is a somewhat dull red color. 4130 only needs a few seconds near 1200 to make the transformation to pearlite, which is what you want---and the rest of the tubing is.

While TIG welding can give very high quality welds, the HAZs next to the welds are where the concerns are---or should be. Gas welding heats a much bigger area/volume of base metal, which slows down cooling thru the critical 1200 deg. range.

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[metengr]

The 4130 is very weldable in this tubing size and would not require any post weld heat treatment. What I would do is use GTAW process with a generous preheat of 300 deg F for the weld joint region.

Do NOT use any torch heating for post weld thermal treatment, you will not need it.

 

[blacksmith37]

Then again you could build like bicycles (were) ; 4130 Q & T ,or normalized, with silver solder socket type joints (that may be the problem). The silver solder does not require a high enough temp (1000 F ?) to soften the 4130.
I suspect the cracking problem with brazing was caused by liquid metal embrittlement (by copper and/or zinc).

 

[CrazyJ]

I would say it depends largely on which sanctioning body you will have to qualify the car with.

It used to be the case that 4130 was used because the rules in SCCA (? maybe NHRA) allowed a thinner wall tube if 4130 was used. I'm not so sure this is the case any longer, but it's worth checking into.

Regarding bicycle construction with 4130 butted with brazes - being that those bicycles were constructed in an assembly line, and the brazes were forged ahead of time, I don't think there are any advantages for the OP.

 

[tensor47]

Hi all.

Thank you all for your views which are greatfully recieved.
It's a long time since I did Eng Materials but I did like this part back then and now the need for understanding it really is upon me.

Can anyone comment on my previous Q1 directly please, as this again confirms, or not, my understanding of what happens.

Q6
This time I have attached 4130 TTT and CTT diagram with what I think are the critical cooling rates so could you please comment directly to this point, as it will confirm or not, my understanding of the digrams.

The diagram is open for edit if you want to add to it or correct it and then repost it as a pic is worth a thousand words to me.

Cheers

John

 

[Tmoose]

Mr Finch is a talented welder, but his comment about 4130's "grain" somehow preventing successful bronze welding/ fillet brazing seems to be his own opinion. Welding near a previously brazed area invites intergranular diffusion by the bronze. Overheating when brazing could do the same thing.

CArroll Smith's "Prepare to Win" has a section on the history and method of fillet brazing race car tube frames. It is completely different than the "lugged" frame construction used on bicycles.
The lower temperature used for bronze welding probably destroys the physical properties less than fusion welding.

Here's a picture of the characteristic brazed fillet joints.
The frame appears to have been plated. I believe Rickman and Seeley were fond of plated frames.

http://farm5.static.flickr.com/4037/4446383932_09ebceb616.jpg

No undercut, a generous fillet, and the stretchier bronze reducing the stress concentration at the fillet toe make for much more fatigue resistant joint than a fusion fillet. But an efficient frame relies on triangulation not weld strength for stiffness.
A roll cage designed with some crushability would be a different story

Tony Foale used fillet brazing on some of his frames.

http://lh6.ggpht.com/_HcrJxB39vv4/S5Bym2C2uXI/AAAAAAAAAqY/HWRCN9xtSIc/s512/Frame06.jpg

 

[tbuelna]

tensor47,

Unless you have a controlled process for achieving normalized properties within the 4130 welds, then to be safe you should assume your 4130 weld zones have annealed properties. Annealed 4130 might have an 80ksi UTS vs. 50ksi UTS for 1020 mild steel.

As others noted, you should analyze your weld joint strengths using the appropriate knock-down factors for weld deficiencies such as porosity, undercuts, thermal strains, etc. Ideally, the weld zones should be lower stressed than the adjacent tubing.

Good luck.
Terry

 

[tensor47]

Hi Chaps,

Thanks for your contributions.

It seems to me :-

The problem with mild steel is that I will loose the extra strength of the cold drawing process when I weld it as it will be back to hot worked condition (yield 51Ksi done to 30Ksi) and there is no way to correct this.
This is what Q1 was about and I never got a definitive answer.
Is there one ?

The beauty of 4130 is it’s good strength (yield 63Ksi) in normalized condition (ferrite & pearlite), and that after welding, if I can get it back to normalized condition, I will have a uniformly strong frame, without the hardness, brittleness or distortion problems of major hardening and tempering.

But is this possible by hand, and that is what Q3 was about and I never got a definitive answer. Is there one ?

If I have interpreted it properly, the CTT shows the max cooling rate to be 50 C/min = 90 F/min between 1450 F and 1000F and the job should be done, a total of 5 minutes. Or is there more to it ?

The thing about nickel bronze brazing, which is how the car in the picture was built, is that it has a melting temp of 910 C = 1670 F (salmon/orange according to Carroll Smith) which is well into the austenizing range, so the normalizing is just as much needed as for torch steel welding or Tig welding with post heating.

Also I don’t think you can actually assemble a square frame with one diagonal using sockets.

Carroll Smith (how much have I learned from this man !) admits pre and post heating may not be necessary, but says it’s good insurance to do it, and I think that’s why some of you advise it and others don’t for this sized tube, so your are all right in a sense. The issue is how near to original normalizing can you get ?

NB
I think the plating on those beautiful Seeley and Rickman frames offers the dreaded hydrogen embrittlement an opportunity, but maybe this was not understood then, as it only became legal problem in Oz in the late ‘60’s.

Cheers

John

 

[Metalguy]

Answers:

Q1--Yes.

Q3--The HAZ'z can be considered to be normalized *if* they don't cool thru the transformation zone so fast that martensite forms down around 4-500 F. You can avoid this by using a gas flame to weld (heats a big area, as stated), or by preheating. Tens of thousands of aircraft engine mounts have been gas welded with excellent results.

While TIG welding can make very nice welds, so can gas, and TIG doesn't heat as big an area, so the HAZs cool down faster.

Remember to periodically examine the weld areas for cracks during service. Slow growing fatigue cracks are always a possibility, but the probability of sudden fast fracture will be very low.



"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[unclesyd]

I talk to a friend that used to work at shop that built racing frames from Indy cars to go carts. He said that they used 4130 exclusively and a few times used CS gussets. All their frames were MIG welded using TIG to for tacking. They did furnace braze some small frames that were cooled to Normalize the material.
There is nothing wrong with gas welding if done done by person well versed in the art. Just remember that all types of steels were gas welded before the advent of TIG.
The welding sire of the day was coat hanger or Ozeld 65.

 

[twalter]

tensor47 -

O/A Welding 4130 is SOP for aircraft frames for experimental aircraft. You can get good repeatable results for home builders. For those welding day in, day out, TIG or MIG is fine. For occasional welding, stick with OA.

One our fellow Bearhawk Builders took OA and TIG classes, and used different rods, but the best part was he recorded all his pull testing and put the video's on Youtube! Nothing an engineer likes more than see actual test results!

http://www.youtube.com/user/alaskabearhawk#g/a

Look at "conclusion" then you can go back and look at his various test. Lots of good material in there.

 

[gadkinsj]

Unclesyd. My experience with welding frames is to use Tig welding for the final welds and Mig welding for the tack welds. I think you got that backwards.

There are places that have furnaces big enough to normalize your frame if you like.

Some form of stress relief is going to happen. You can choose to have it happen from the vibrations induced during racing, or get it done after you weld it so you know it is aligned properly before you start bolting things to it.

 

[Mr168]

Just my $.02 weighing in, but in scenarios like this, it's often just as important, if not more so, to consider real-world results as it is the theory supporting them.

If the same size material will be chosen regardless of alloy, then the discussion should end right there. Without the weight advantage of a lighter walled 4130, mild steel takes the nod.

I've seen many a wrecked chassis, and I have never seen a structural weld failure on a mild steel car that I couldn't immediately identify as the result of a poor weld, whether it be an obvious stress riser, porosity, or most commonly lack of fusion/depth of penetration. On the other hand, I have seen many weld failures in 4130 cars, both down the weld centerline (often the result of undersized, concave fillets due to the "heat input" paranoia,) and at the toe of the weld. And in a number of these failures, the vehicle hadn't even seen a wreck yet.

Normalizing a 4130 assembly without significant distortion is not a cheap or easy thing to achieve. To be honest, I'd rather see a mounting tab or tube bend after a light crash, than I would a straight piece of tubing or tab with a hairline crack at the toe/HAZ that I missed. Given the very short runs of tubing and triangulation in a rear space frame, I'd be very surprised if you saw any advantage, crash or otherwise, other than the one seen on paper.

 

[Metalguy]

>"On the other hand, I have seen many weld failures in 4130 cars, both down the weld centerline (often the result of undersized, concave fillets due to the "heat input" paranoia,) and at the toe of the weld. And in a number of these failures, the vehicle hadn't even seen a wreck yet."<

"Heat input paranoia"? That's because the welder has it backwards. With something like 4130, more heat is better than less. Ever heard about untempered martensite?

Next, "toe of weld", as in HAZ. See above sentence. Race cars and airplanes are similar--both need high strength-to-weight, and mild steel isn't where it's at.

"You see, wire telegraph is like a very long cat. You pull his tail in New York and his head is meowing in Los Angeles. Do you understand this? Radio operates the same way: You send signals here, they receive them there. The only difference is there is no cat." A. Einstein

 

[swall]

I used to work in the heavy equipment industry and we made weldments consisting of plate, tube and steel castings. The castings were at the stress points. Although everything was done in a fixture, TIG was used to tack everything together and then MIG used to lay down most of the weld metal. The idea was that a skilled welder was needed to put things together, but once that was done, anyone could lay the metal down with MIG.

 

[Mr168]

Metalguy: You're preaching to the choir in regards to heat input.

There's a significant difference between what welders should do, and what they actually do, especially in race fab shops and garages, where appearance is everything, fillet size pertains only to a steak dinner, and the pencil-thin TIG bead reigns supreme. I've also seen many people try to gas weld thin-walled tubing who couldn't set a neutral flame if their life depended on it.

If the welder skill and knowledge is there, go 4130. But once again, based on my personal experiences, 4/5 vehicles I've seen TIG welded with 4130 base metals were bad enough to give me goosebumps, and the failures I've seen were as predictable as the sunrise. I've never encountered a gas welded automotive chassis, so I can't comment there (those I knew who have attempted it on small aircraft had years of experience, so I can't lump them into the realm of "garage warrior").

I'm not convinced that a lightweight, relatively low-powered hill climb car NEEDS 4130. I'd think the fact that the OP has a .0625 walled mild steel cradle on it's 43rd year of use is a testament to that.

 

[unclesyd]

gadkinsj,
i rechecked my information a I had it right, no senior moment. They tack with TIG and weld out with MIG. I alos asked about distortsion during any heat treatment and his response being that it never gave them any problems.

Addenda:
Jimmy just got a pacemaker and can't get within 40' of a weld cable after having welded and taught welding for 43 years. he was one of the better and fastest welders in the area.