Why did this metal fail?

auditdata

auditdata

Well-Known Member
Joined
28 Oct 2009
Messages
303
Location
Chester
www.sephina.org
This is the davit base from a friends yacht. It is less than a year old. The davits are rated at 60Kg each. The dingy is 52 kg and the motor 25kg. So the load is within spec. The top two bolds are welded in where as the other 4 are not. Presumably this was to aid mounting, as the angle would make insertion difficult when fitted. It surprises me that whatever load made the two top bolts fail it did not make the lower ones fail. The plate was flat originally. Is this a failure because of the weld?
Any advice and should the new one not be welded? Am I on the wrong track?
Pics below
DSC02036.JPG

DSC02034.JPG

DSC02053.JPG

DSC02049.JPG

DSC02057.JPG
 
Last edited:
The 60Kg is the total load for both davits or one davit individually. As there is a heavy motor on the dinghy all or most of the weight is taken by the davit nearest the engine. Therefore davits overloaded and failed.
 
I would hazard a guess that the bottom one (first pic) has failed in torsion (the radial step and helical fracture face). Maybe the nut 'picked up' when it was being tightened, or perhaps was just over tightened.

The top one looks like it has failed due to fatigue. The photo isn't really clear enough to make out where the crack originated. The fact that the section at the fracture isn't completely round means that the weld has undercut here. Don't know whether the failure was due to overloading due to the failure of the 1st fastener, or just poor design. Welding a threaded section is just asking for trouble.
 
auditdata said:
60kg Each so 120kg in total.
Click to expand...

That's with both davits equally loaded.

With an outboard on the transom of the dinghy, the centre of gravity of the outboard and dinghy combined with be much, much closer to the transom than the mid point of the dinghy. So the davit nearest the transom will be taking the vast majority of the load. In fact, there is probably a 'single point lift' possible with only that davit.

Add some rainwater or spray and the load will be way over 60kg . That davit should be rated at the full load plus a safety margin.

The failure has occurred at the boundary of the heat affected zone where the captive bolt has been welded in. As this is an inevitable result of the manufacturing process, it will have been allowed when calculating the 60kg SWL. (It would be interesting to know what the allowed safety margin is, but I would think is was very minimal - hence the need to understand the loading characteristics of the dinghy).
 
Last edited:
I'm no welding expert, but my guess is that wasn't so much the weld as the post weld heat treatment, which I guess was none.
The bolt will have got hot enough to change the crystal structure of the SS, which will have been "frozen" by rapid cooling, making it the SS very brittle and the crack will have progressed from the root of the thread.
There has been extensive research on how to weld SS, I don't know the details, but when welding SS you need to be very careful about the pre-weld heat treatment, the filler rod, the type of welding equipment and the post weld heat treatment.
Most things in the sailing world are over engineered and not that critical, so we get away with poor weld procedures most of the time.
 
PCUK said:
Yes, and all the weight is on the engine end davit. so overloaded. This is simple physics and the way davits are rated. 120Kg only applies when the weight is evenly distributed on both davits and in this case it is not.

You beat me to it Tim!
Click to expand...
quite,
optimistically the engine end davit is taking engine + half the dinghy ie 51kg.

But realistically the dinghy is wedge shaped so the engine end will be more than half the dinghy.
Then there ar dynamic loads from bouncing on waves etc,fuel load....

The engine davit will certainly see more than 60kg at times.
 
Not the easiest photographs to interpret but the fillet in the top pic appears to have failed within, or very close to, the weld. Weld quality in each photo is poor. Assuming austenitic stainless steel for the plate and the bolt, then heating to weld temperatures will not transform its metallurgical condition and neither pre nor post weld heat treatment would be needed for a simple job such as this one. The lower pic is too out of focus to diagnose but there may be evidence of multiple origin fatigue. As already pointed out, welding a bolt in this way is bad practice, particularly done as poorly as this. Whether a well- made one would have survived seems questionable with a load close to maximum at such an acute angle (if I am interpreting the design correctly).
 
Why is everyone going on about weld failure? The engine end davit was overloaded and that's that. Ironically, if another 25Kg engine had been fitted at the bow then this would not have happened as the weight would have been evenly distributed. Don't impugn the davit makers name for the lack of understanding of the boat owner.
 
chewi said:
quite,
optimistically the engine end davit is taking engine + half the dinghy ie 51kg.

But realistically the dinghy is wedge shaped so the engine end will be more than half the dinghy.
Then there ar dynamic loads from bouncing on waves etc,fuel load....

The engine davit will certainly see more than 60kg at times.
Click to expand...
The engine davit will take all the weight all the time plus stress loads at sea.
 
PCUK said:
Why is everyone going on about weld failure? The engine end davit was overloaded and that's that. Ironically, if another 25Kg engine had been fitted at the bow then this would not have happened as the weight would have been evenly distributed. Don't impugn the davit makers name for the lack of understanding of the boat owner.
Click to expand...

That would be the case if the bolt failures were simple cup and cone fractures, which would indicate a ductile failure due to overload. Although it is difficult to say just what they are due to the poor photography, they are definitely NOT cup and cone. Nobody can doubt that this is not top quality welding.
 
misterg said:
Welding a threaded section is just asking for trouble.
Click to expand...

+1
One of the aims when welding is to produce a smooth fillet with no sharp angles or crevices that will concentrate stresses. This is almost impossible to do on threaded sections.
Looking again at the photo, it appears that the bolt has been welded on the underside. A highly skilled welder may have been able to pull this off as compared to the amateurish welds shown. Even then the metallurgical changes from welding stainless(Carbon precipitation in the heat zone, plus grain growth )mean that it was doomed from the start.
If the aim was to lock the bolt in place a simple tack on the outside between the hex head and the base would have been better. Executed by a professional welder this weld need not be big and unsightly.

Even without the weld though, alternating bending loads on threaded sections of stainless steel will always produce a fatigue failure. The time to failure can be calculated from the amount of bending stress x the number of cycles.
With the bolt head welded, and the shank trapped in the hole, any slight movement of the stanchion will translate as bending stress on the bolt.

Good luck
 
Overloaded but a horrible design in the first place. I'd bet a couple of quid that the rust on the welded on tube is emanating from a crack at it's junction with the weld. The screws (welded because it would have been difficult to get a screw in once the v was welded?) just happened to beat it to failure.
 
The bad weld is the cause of the initial failure. From a structural engineering point of view an assembly rated for a 60kg dynamic load should have a safety factor of at least 150%. In short, plastic deformation of the structure should only occur at loads of 150kg. The catastrophic failure ought not to have occurred until plastic deformation had already begun to take place. This would not appear to have happened here.

I have tested welds in tension and bending. A good weld should be at least as strong as the surrounding metal.

The photograph does not show much, but I would have expected to see some tearing of the weldmetal and the base plate if even a small part of the weld had been done properly.
 
Last edited:
The only way I would do something like this with studding is either to tack weld a nut to it on the top face and if necessary lock the nut to the top face with a tack.Its then the nut on the thread which tales the load not the weld.
The amount of grind off on the bottom of the plate suggests a poor weld which I recognise as an amateur welder who sometimes doesnt get it right!Similarly looks like some of the studdings diameter has been cut away by the welding.
I find working with 2.5mm stick rods at 60 amps I have to move the rod pretty fast to get a reasonable deep weld on 3mm thick stainless and welding tubing it only takes a moments hesitation to burn a hole.
However I agree that to bend the plate means that the construction isnt strong enough for the load its carrying.
 
At 50 kilos this is quite a dingy and 25 for outboard yelds it is not 3.5Hp.
A picture of the davit in place would help but as others pointed out it looks flimsy for such a load.
 
Last edited:
You must log in or register to reply here.

Share this page

Top