Welding the skin to the frames demystified

[M&M Ovenden]

According to ABS, the skin of a hull should be staggered welded every 12 inch with beads of length relative to the skin thickness. The minimum rules are clear. I'd like to get your word on over welding the skin to the framing. What are the consequences, is there any benefit?
My understanding is that more welding does make a hull stiffer but also increase the chances of obtaining a hungry horse shell, even more so on more flat paneled constructions. Does more welds make a boat stronger or does it increase internal constraints and stresses to the hull. Are more welds only negative on the work load and aesthetics or does it have negative impact on the strength of a small vessel? How much is too much. I have done my reading about all this and could spit it all out, what I'd really like is an experienced input.

Thanks,

Murielle

 

[Brent Swain]

The best way is to skip the transverse framing altogether in boats under 55 feet. They are structurally irellevant.If you already have a framed boat , the only way to avoid the hungry horse distortion is to leave the stringers free in the frames, unattached while you do the longitudinal welding, and stringer welds. The shrinkage from this will lift the striingers off the frames an inch or so forming a slight compound curve in your topsides etc. Then you weld only your stringers to the frames wherever they happen to be.
Hungry horse effect comes from shrinking the edges of the plate while the plate is attached to the frames, and can't go anywhere when the edges shrink. The shrinkage then pulls the edges of the plate shorter , leaving excess plate between the frames and stringers.
Brent

 

[MikeJohns]

Murielle

A good question

If you used fabricated T bar frames then you have already made the first call on this when you welded on the flange.

Shrinkage (with consistent welding), is also consistent, predictable and measurable, you can always run up a few test pieces if you still have scrap around then simply flat edge and feeler gauge measure for comparison after tacking then after welding for different welding patterns.

Overzealous throat thickness/penetration seems to be the great distorter of panels at this stage , sticking just above the scantling throat minimum and using a MIG wisely is going to be OK to reduce the spacing . I also prefer a closer spacing than our own USL standard particularly the fwd 1/3 of the vessel 70-75% on chain and staggered welds as it definitely adds to the strength for more severe loading.

In severe grounding damage I have often seen the intermittent weld either broken or torn through the plate. Also of interest here; one line of trawlers with bow flare had problems fatiguing and breaking intermittent welds forward in heavy weather, we continuous re-welded the forepeak frames to the plating and there were no further problems. To give an idea of the heat the outer existing paint ( black epoxy ) only just discolored and was left intact with an overcoat but that was (from memory) 8 mm welded up low and slow by an experienced fabricator with zero visible distortion .

There is more to it than just holding the plating on as for example with a fastened wooden vessel, in steel scantlings rules the plate is counted as structurally part of the T or I beams that the transverse and longitudinal stiffeners become. They are calculated with the plating "attached" . If your framing is heavier than the rules require then you are in credit.

Continuous welds are demanded for high strength applications, military vessels submersible craft etc because it adds a lot to ultimate strength once any part of the structure gets past yield. Scantling rules work on yield plus a safety factor they are not concerned with ultimate strength. This is where you can add some insurance, but if you have very thin hull plate it is debatable if it is worthwhile.

Note that you tend to get more distortion in areas where you have a higher pre-stress (plate pulled in) usually as visible but shallow dimples. Sometimes when fully welding these panels will dish more and are flattened out again with external line heating but the point is even fully welding with care and they don’t dish noticeably.

You know most of this anyway.

cheers

 

[M&M Ovenden]

Yes, I do know most of this but there is a difference between the theory one reads, ones experiences and conclusions coming from interrogations and discussions with others.
Mike, you just made a blurry image quite clearer. I get a boat can't really be structurally over welded (which is were I was really unsure in the case of smaller hulls), a stiffer hull only gets stronger, but there is no point having too much throat if there's enough bead. Right?
I have finished welding up my hull last year with good results and a lot learning but I can't stop the questioning, boats are captivating structures.
I have seen many older boats with the skin not continuously against it's framing, probably built on floating stringers as mentioned by Brent. Wouldn't a boat built such a way tend to fatigue its frame to skin welds more? Seems like this days with much more precise frames (CNC cut/ CAD/ no lofting) and better controlled welding equipment there would be no need to have floating longitudinals.

Thanks,
Murielle

 

[Wynand N]

Frames do not need to touch the hull skin - better to have "floating" transverse frames and the hull plating attached to longitudinally stringers only. This results in extremely fair hulls as my boats are testament to and the "hungry horse" look eliminated.
All my steel designs are designed this way and so all the later Dix boats I had built, the Dix 65, Dix 57, Dix 43 and Dix 38, and all these design exceeds ABS scantling rules.

I attach two photos showing this and it can be seen that the hull plate clears the frames by about 20mm except in high load areas such as keel, etc where the frames actually are against the hull plating, which incidentally, are under the waterline and not noticeable if some minor distortion were caused by welding...

 

[M&M Ovenden]

That does look very neatly built Wynand. Looks like your boat frame is initially built with the longs evenly proud of the frames. It doesn't seem like you adjust the longitudinals to the skin (if you do t must be minimal) but rather that you tie your skin to a precisely built frame. Is that right?

 

[Brent Swain]

The amount of lifting of the hull off the frame is clearly shown in your first photo. This is what steel wants to do when you weld the longitudinal welds. Trying to stop it by welding everything solid to the frame before doing longitudinal welding is the main cause of the hungry horse look.
Longitudinals are a must to avoid the hungry horse look. The same applies to side decks.
How big a boat are you building? Hard chine or round? How thick the plate?
Brent

 

[Wynand N]

yes and no.
Here is my procedure; the frames get erected upside down on a strongback and the stringers dropped into there designated slots cut before the frames were set up. At this stage the stringers stand out proud of about 20mm and as the are put in place, the easily follows the curve of the hull and are very lightly tack welded to frames.

Then the plating goes on and the plate allowed to take its own natural curvature to prevent flatspots and tacked together and to the stringers where the plates and stringers are in contact. When the plating is done, one goes under (inside) the hull and start fairing the hull so to speak. If there is a flatspot on the plating, I knock the stringers loose from the frame at that spot and drive a spike in behind the stringer to the frame and push the flat out with the stringers - if fair, lock the stringers by welding the stringers to frame. Where the stringers are a bit away of the hull plate, one simply push the stringer against the plating and tack weld. When all is fair and smooth, welding can start.
As you had said, if adjustments were to be made, it is indeed minor ones.

Hope this senseless rant made some sense

 

[Wynand N]

How big a boat are you building? Hard chine or round? How thick the plate?
Brent

Brent, she is a Dix 43 (43ft 10") radius chine hull. Hull plating from sheer to keel is 4mm and the keel sides 6mm plate. The deck is 3mm all over.

Here are some photos - my apologies to others that have seen them already - of her and you can see what a fair hull this is. The photos was taken just after the epoxy primer was applied and no filler used at this stage.

 

[MikeJohns]

.......... I get a boat can't really be structurally over welded (which is were I was really unsure in the case of smaller hulls), a stiffer hull only gets stronger, but there is no point having too much throat if there's enough bead. Right? ...............

Yes on both counts.

Wynand's boat there is an interesting example (or rather masterpice of the art) with such thin plating and continuosly welded floors he still has no visible distortion on the garboard.

Wynand
That would be ABS racing craft scantlings presumably? I remember trying to bite you a long while back about such thin plating, 3mm decks ! ...... go fast boats.

 

[LyndonJ]

Frames do not need to touch the hull skin - better to have "floating" transverse frames and the hull plating attached to longitudinally stringers only. .............

Query
If the transverse frames Section Modulus is calculated with the skin plate attached you are weakening the transverses considerably doing this. Then the framing needs to be heavier to compensate. So you cant just take the ABS scantling frame calced with attached plating and then offset it.

If you do this it should be specifiacally designed for, ....be careful.:!:

 

[Wynand N]

Query
So you cant just take the ABS scantling frame calced with attached plating and then offset it.
If you do this it should be specifiacally designed for, ....be careful.:!:

For starters, all of the Dix designs I had built, were designed with floating frames and excess of ABS scantlings per say. The Dix 57 I had built, the client actually insisted on ABS approval and the plans passed without any alterations.
Secondly, as a designer myself, I would never change anything structural without the written approval and blessing of the designer involved - regardless...

Mike, the Dix 65 I built in 1990/1 was also build with 4mm plating from sheer to keel with the keel sides 8mm plate. Again the deck was 3mm. And she is still sailing the world. It shows what a word class designer can do with proper engineering principles applied designing a strong light steel boat.

 

[Brent Swain]

Murriele
What kind of boat are you building?
John Dearborne ,who builds beautiful round bilged aluminium boats in Gibson's BC , said that if the transverse frames even touch an aluminium skin, even without welding them to the skin ,the frame can still be seen showing thru on the outside of the boat, after the welding is done.
Over welding is not a structural matter , its a question of doing far more than is needed and how much resulting distortion you can accept.
Sure wish I could get that 4mm here in Canada for my 31 footers. We went metric for most things , but not steel sizes . Having access to both would be ideal.
Brent

 

[M&M Ovenden]

Brent,
You can see my boat on my web site. It is a steel reproduction of a traditional Brittany tuna fishing dundee, the hull is 50 ft and will displace a heavy 55000 lbs.
On our previous boat, we have taken the bad habit of winterizing in the ice, which can explain the particularly heavy scantlings of our future vessel.

 

[MikeJohns]

Floating frames would certainly make fairness easier with light plating but.....

As pointed out above the frame calculation is not trivial, it is interesting because by moving one face of the transverse frame (the hull plate) farther away, we increase the section modulus but at the same time introduce some undesirable stress concentrations and change the loading patterns considerably.

The stresses run from the plate through the stringer transverse contact and this contact is over a small area, consequently the transverses will have quite different stress patterns and higher stresses in both the top and bottom of the section. The longitudinal will also be under a higher stress immediately under and adjacent to the transverse. This stress will need to be designed for . All the load is transferred via the point contacts which is not all that desirable from a structural perspective.

Note this standoff method would suit flat bar better rather than a T section. A face bar moves the neutral axis and increase the stress in the bottom of the resultant T .. If the stem of the T is thin then it is not good design practice to highly stress a thin edge..

From a quick look at this the stresses with be around 25% more in the high load areas created. This stress increase will occur as a localized load in the stringer and more global in the transverse. It is possible with this construction in steel that you will see some fatigue issues after prolonged heavy weather if the boat was driven hard with high slamming loads. In aluminium alloys I would guarantee it unless very close attention to design and construction detail was maintained.

Grounding damage will be more extensive since the ultimate failure mode of the framing will be much earlier buckling (as opposed to yielding) without the restraining effect of the attached plate.

The compromise is ease of building and fairness against the ultimate strength that could be achieved with the amount of material used.

If anyone is really interested in this it could be illustrated easily enough.


A note on ABS

ABS offshore racing craft is a (now dead) self compliance standard it has been shown to be deficient in way of deck loads ( not modified) and GRP keel attachments (modified ) and it is no longer being assessed and re-written, neither is there any approval process, just statements of self compliance from designers.
Isn't the advice from ABS themselves for designers to to adopt one of the ongoing small vessel standards eg DNV, GL, ISO, Lloyds.
Scantling rules need to be modified on an ongoing basis as and if they are shown deficient. Vessels going into survey need a full approval process; plans and construction supervision and then you cannot use ABS themselves but the designer can claim compliance .

Some of these up to date scantling rules even embrace “frameless” construction..