Testing your welds !!

[Bonzo]

I have been meaning to make a post on this topic for a very long time.

Been a nice day today and having nothing pressing to do, I thought it was time that I did.

This post is very relevant to a recent thread elsewhere but I guess this is the correct place for this post as it is mostly relevant to building the chassis.

Spent an hour or two this afternoon making a few welding test pieces.

The test pices were 25mm X 25mm, 16 gauge ( 1.6mm ) ERW square tube off cuts.

The welder used was a Lincoln Powertech 191C ( 180 amp ) .8mm wire & sheilding gas was, BOC Argoshield Universal

All of the test welds were un-prepped & but welded with no weld gap.
None of the material had been subjected to heat & all welds were started from cold.

Once the test sections were welded, the welds were cut out from the square tubes to produce, flat test strips.

Here is a sequence of photographs.



For ease of identification the test pieces are marked A, B & C

Looking at weld A you may think it looks reasonable.
Nice narrow weld bead that will not take too much work to fettle & make flat.
However, looking at it with a trained eye, allarm bells start to ring & the weld looks cold.

Weld A used a low current & wire speed setting.


Weld B looks fine & used a sligtly higer current setting with increased wire spped

Weld C looks fine with exception of visual evidence of being a tiny bit on the hot side !!??
This was performed with another increase in current & wire speed.

Now we flip the samples over to reveal the truth.



Instantly it is evident that weld A lacks any form of root penetration.

Weld B has a full root penetration ( A tad shallow )

Weld C has a good root penetration ( A tad too much )

To be ideal, a weld setting in between B & C would produce an ideal weld, one that I would be happy to submit to any form of testing.

That said, in my humble view, either weld B or C is more than suitable for the Roadster chassis & would easily pass BS standards testing.

More to follow in part 2 , just as soon as it is typed in a few moments from now

[Bonzo]

Here we go for part 2 of this post.

Here is a picture of welds A, B & C after the weld cap has been removed ( Fettled flat )



On close visual inspection of all 3 welds, from the outside face of each joint there is nothing visualy to distinguish any one from another.
No porosity evident, no inclusions evident.

Without polishing & etching, all welds appear to look as they should do when making a simple visual inspection of a weld.

Now here is a picture of wels A, B & C after they have been subjected to a very mild bend test of about 100 - 120 degrees
A true test would be made by bending the sample through a full 180 degrees, with a radius appropiate to the thickness of the material being tested !!



It is self evident that sample A has failed the bend test miserably & is a totally unsuitable weld joint for the Roadster

As there was really nothing to choose from between the next 2 samples, I chose to test the strength of the root weld sample B ( Could be argued that it looked the weaker of the 2 remaining samples )

Tested sample B buy bending pressing the dressed face to put maximum stress on the welds root.
The root held firm & I am confident that it would bend to 180 degrees without showing any signs of stress fractures.

Sample B was tested in the other direction to test the front face of the weld.
Nothing undue was observed.

In sumary, if you are unsure of your welds, keep practicing & don't be afraid to test your welds frequently .... It won't take long & will give you peace of mind, knowing that you will be building a safe, sturdy chassis

I hope this helps some of the novice welders on the forum.

[eSteve]

Nice work Bonzo, thanks for putting in the effort and posting it up

[HandyAndy]

Ronnie,

Thats a really useful thread , very informative & with pictures that actually shows what we,re looking for in a good weld ( a picture tells a thousand words so to speak ).

Nice work Ronnie

cheers
andy

[Enoch]

What a useful and informative thread. very well put together and very well executed. I salute you, sir
Enoch

[Talonmotorsport]

Very nice work there Ronnie, perhaps you could put what rough amps you used to weld A,B and C.

[Bonzo]

I thought it best that I completed a full bend of the test pieces

Here is a picture of the fully bent pieces.



Test piece A folded in half by hand !!, if I were to try & straighten it again, one piece will become two again.

The other two samples exhibit absolutely no signs of stress fracture, if test piece B looks a bit iffy, I can assure you that it is only a bit of mill scale as the test pieces have not been cleaned/polished

Basic testing of a weld sample is quite simple to do at home, well for thin material anyways.

If you do not have a press, a simple vice and an appropiate piece of round bar will do the job fine.

Open the vice a little, place the piece of bar directly over the centre of the weld on the test piece & strike with a suitable hammer.
This will allow you to bend the test piece a fair way.
To complete the bend, just pinch the piece in the vice until fully bent, when I say fully bent, I don't mean squished flat

Once bent, examine the surface of the welded area for signs of stress fracture.
If you want to examine in more detail, fettle & polish the weld with emmery paper, this will help to highlight any imperfections in the test piece

I have not gone into fine detail of the weld settings used for the test pieces because.

A: My welder does not have any digital read out of the weld current.

B: The settings will vary between welders, even identical machines !!

Hopefuly I have a least covered the basics of testing your welds.

Setting up the welder & weld prep is another story waiting to be told.