Partial Bulkhead Scantling Calculation ISO 12215

Discussion in 'Boat Design' started by ToMeK, Oct 1, 2021.

1. Joined: Nov 2006
Posts: 42
Likes: 3, Points: 8, Legacy Rep: 10
Location: Zagreb, Croatia, Europe

ToMeKYoung naval architect

Dear Boatdesigners,

I am kindly asking for an advice someone experienced with ISO 12215-5.

In the project I am working on at the moment it was decided to use plywood partial bulkheads as stffeners for Hull bottom and at the same time suport for Module line (as can be seen in the picture)

Plywood partial bulkheads will be glued to the hull and overlaminated, then Hull liner will be glued to the patial bulkhead.

My question is, how to calculate those partial bulkheads according to ISO 12215-5?
I have seen some recommendations to calculate them as stiffeners, but since they are not consistent in height over some logical length (l/b > 5) I am not sure whether this is correct.
Another suggestion was to caluclate it as watertight bulkhead, but it is not watertight.

Finally I came across Caluse 9.3.4. Non- watertight or partial bulkheads; referring to anex A.13. It says that solid plywood partial bulkhead can be calculated using formula tb = 7,0 * Db.
Could this be the best way to calculate it? In my case local heights from keel to sheer are about 2m giving bulkhead thickness around tb = 15mm; which seems not very strong to cope with slamming loadings on the bottom.

Can anyone please comment on this.

Sketch of typical "Partial bulkhead" is attached.

Many thanks,

Kind regards,

Tom

2. Joined: Nov 2006
Posts: 42
Likes: 3, Points: 8, Legacy Rep: 10
Location: Zagreb, Croatia, Europe

ToMeKYoung naval architect

In some section partial Bkhd looks like this

3. Joined: Sep 2011
Posts: 7,458
Likes: 729, Points: 123, Legacy Rep: 300
Location: Spain

TANSLSenior Member

The model in the attached figure is intended to suggest that these "bulkheads" (floors) could be calculated as if they were stiffeners subjected to the loads on the bottom or on the liner.

4. Joined: Nov 2006
Posts: 42
Likes: 3, Points: 8, Legacy Rep: 10
Location: Zagreb, Croatia, Europe

ToMeKYoung naval architect

Exactly, but then the positive influence of liner giving support to stiffener (and hull as a consequence) is left out.

5. Joined: Sep 2011
Posts: 7,458
Likes: 729, Points: 123, Legacy Rep: 300
Location: Spain

TANSLSenior Member

No, because you consider the floor as two independent stiffeners (two flat bars), the upper one with the corresponding attached liner plate and the lower one with the bottom attached plate. They are calculated as a normal deck cross beam and as a normal bottom frame. The width of the attached plates will be given by ISO 12215-5.

6. Joined: Nov 2006
Posts: 42
Likes: 3, Points: 8, Legacy Rep: 10
Location: Zagreb, Croatia, Europe

ToMeKYoung naval architect

I agree with that, but what about the beam rule that l/h must be bigger then 5?

7. Joined: Sep 2011
Posts: 7,458
Likes: 729, Points: 123, Legacy Rep: 300
Location: Spain

TANSLSenior Member

Sure, any reinforcement, real or simulated, should comply with the rules regarding reinforcement. But I think l/h> 5 cannot be correct.
It is very important to take into account the position of the plywood layers in relation to the loads that it will bear. You must calculate them as edge plywood.

8. Joined: Nov 2006
Posts: 42
Likes: 3, Points: 8, Legacy Rep: 10
Location: Zagreb, Croatia, Europe

ToMeKYoung naval architect

Then I must have mixed something. Probably with classic beam theory.
Of course, this is very important.

Thank you for your help TANSL.

9. Joined: Oct 2008
Posts: 7,828
Likes: 1,709, Points: 113, Legacy Rep: 2488
Location: Japan

ToMeK

I can see where you are having problems in understanding what to do.
The first issue you need to decide, when designing structure, is to decide which structure is supporting what?!

One structural member supports another. In doing so, there ends up a primary structure and a secondary structure. The primary structure being the main supporting element.
This can be either transverse or longitudinal... the choice is yours, it is not a rule thing, per se.

In the example below:

We can see the basic structural arrangement. In this case, the frames, the transverse frames, are the primary structure.
Their spans go from major discontinuities to another, such as from chine to keel, like so:

The discontinues are usually major changes in hull section shape.

Thus, once you understand this, you can make the vessel all transversely framed and no longitudinals, or, all longitudinally framed, and no transverse frames....as such, the choice is entirely up to you. ...but the convention is usually transverse frames - the primary structure, as noted by the spans shown above...and these then support the secondary structure, the longitudinals. The spans of the long.ts are then between each frame, shown by the first image.

So, to analyse your structure is the vessel transversely framed with long.ts spaced at "some distance" apart, shown in fig 4-45?
If this is the case, your "partial bulkhead" is just another frame. As such the span is from:

But, the problem you now face is that, the frame using this span, and the frame spacing, will most likley show that these regions here:

Are too weak and fail. Because if you take a section, on the blue line, the modulus at this location will be less than that shown by the black like:

It is highly likely that the modulus required at the blue line will mean a major change in structure that would not be suitable for your arrangement.
Solution..

Again make this section of the frame (partial bulkhead) the secondary structure, so it is then supported by a primary structure, like so:

The blue lines are longitudinals that become the primary structure. And depending upon how you arrange the entire vessel these can span just between the frames that you need, or the length of the boat to suit.
Since the span of the frame is now reduced to thus:

Meaning the minimum section modulus can most likely be satisfied.

Does this make sense?

Last edited: Oct 3, 2021
rxcomposite and bajansailor like this.
10. Joined: Oct 2007
Posts: 3,649
Likes: 1,593, Points: 113, Legacy Rep: 37

bajansailorMarine Surveyor

Absolutely!
(to me at least it does).
What an excellent way of explaining the importance of (and difference between) primary and secondary structures.
Primary longitudinal framing here certainly seems to be the way to go.

11. Joined: Nov 2006
Posts: 42
Likes: 3, Points: 8, Legacy Rep: 10
Location: Zagreb, Croatia, Europe

ToMeKYoung naval architect

Thank you for your very long and detailed answer. My mistake is that I haven't drawn in longitudinal structure. Of course, hierarchy exists. Boat is transversally framed with longitudinals being secondary. Some frames are GRP, some are overlaminated plywood (like the first sketch). But in some places, where frames were spaced to far away and where slamming loads occur, intercostal stiffeners are added transversaly (second sketch).
I am not quite sure whether module liner sides in longitudinal direction can be assumed as higher order structure because the are not continuous. They appear only locally and cannot be assumed to be supported by other liner surfaces.

My question is more in direction to find out what "partial bulkhead" in ISO 12215 means and where can this formula of tb=7*Db be applied. I have also tried using other methods like Gerrs Sn method, and according to his calculation plywood bulkhead thickness comes up similar to this equasion.

12. Joined: Nov 2006
Posts: 42
Likes: 3, Points: 8, Legacy Rep: 10
Location: Zagreb, Croatia, Europe

ToMeKYoung naval architect

Mixed type of framing has been employed here. Longitudinals are strong, but they span in between four strong frames or bkhds: Collison Bkhd, ER Bkhd, strong GRP frame thats in the middle and one plywood bkhd infront. Other structural members are transverse plywood and are intercostal.

I know it is a bit confusing concept, but I joined the project in quite late stage where all other space reservations were made already but structure, so we were looking for solution that demands least changes in other arrangement.

Attached is better sketched forward plywood bkhd that is supporting GRP longs

Attached Files:

File size:
104.2 KB
Views:
139
13. Joined: Oct 2008
Posts: 7,828
Likes: 1,709, Points: 113, Legacy Rep: 2488
Location: Japan

Partial means, as per ISO definition.
Basically it just means bulkhead that may not be full height. That being from keel to deck.

The tb=7.Db, is simply a reflection of this.

It just means that if you have a partial (or otherwise) bulkhead that is say 1.0m deep, the thickness must not be less than 7x1.0 = 7mm.
If the partial bulkhead is say 1.5m deep the thickness is not less than 1.5 x 7 = 10.5mm.

The key part is this: " ...The thickness of unstiffened solid plywood bulkheads..."

So if you are using plywood..and if the bulkhead has no stiffening on it (for whatever reason) the thickness must not be less than calculated by said formula, in 11.8.1.
That's it.

14. Joined: Sep 2011
Posts: 7,458
Likes: 729, Points: 123, Legacy Rep: 300
Location: Spain

TANSLSenior Member

What appears to be inferred from ISO Chapter 9.3.4 is that a partial bulkhead is a structural but not watertight bulkhead. Accordingly, the bulkhead should be calculated as a reinforcement. Therefore, the formula tb = 7.0 * Db cannot be applied to it, nor can any of the formulas in Table A.13.
By the way Db is :

Regarding what is said in 11.8.1, I cannot comment because in my copy of ISO 12215-5 I have not found that chapter. Sorry.

Another very important thing to consider when calculating any element of a structure is how that element is working, and what loads and how loads are applied to it. Put simply, a bulkhead has loads perpendicular to its plane with a trapezoidal distribution, as a function of the height to the top of the bulkhead. A panel also works in a similar way, except for roof panels which have a more uniform load distribution. Both are calculated according to the thin plate theory, for a plate with its four recessed sides. A stiffener is calculated according to the theory of beams, with loads perpendicular to its longest dimension, subjected to bending and with both ends embedded. A pillar is calculated in compression, for example according to the Euler formula. Therefore, whatever the name we give to an element, its calculation will be made according to how it is working to help the resistance of the structure.
After analyzing all the above, in my opinion, the partial, non-structural bulkhead that the OP shows us should be calculated as a stiffener.

15. Joined: Nov 2006
Posts: 42
Likes: 3, Points: 8, Legacy Rep: 10
Location: Zagreb, Croatia, Europe