Large Rectangular Door in Ellips Head 1

[Crompto29]

I am designing a 1800mm high by 800mm wide doorway of 80mm thickness into a 2300ID Ellipsoidal head. The Pressure is 6 barg and material is A516 Gr70.

See Attached Drawing.

I have designed the Door way so that it will hold up to all the loads on it before it is even welded into the Head.

I have carried out regular compensation calcs on the Head assuming the doorway is a round nozzle of 1940mm ID. This is the distance from corner to corner of the doorways radiused corners.

Has anyone out there had any experience with this sort of design?

I ahave had a look at just puttin gthe doorway in a flat head but head thickness comes back as over 140mm. To thick.

 

[alanw7272]

Crompto29,

May I ask why? Why not use a standard circular manway?
I have only seen rectangular openings in the shell section before, not a dished head and especially not the knuckle region. Also you cannot simply treat this as a circular opening to Code Rules. You are going to have high stress intensities at the corners and will in all likelyhood require FE Analysis!

Perhaps explain what you are trying to achieve and others might have experience?

Cheers

 

[Crompto29]

Thank you alanw7272,

The vessel is a Hyperbaric Chamber for a Hospital. The Nurses need to roll a hospital bed into the chamber without having to lift it through a round manway. There are also physically impaired patients that are unable to climb through a round manway.

These rectangular manways are used in Chambers quite often. I agree that the doorway should be at least kept within the knuckle tangent and probably will be in the final design.

I'm querying anyone that might have had experience with something like this. Or could direct me to any refences.

Ideally I would use FEA, if I could only get access to it.

My third party regulation orthority (in the UK) believes my method above is acceptable. (It wont be U-Stamped.)

My final design may need to be verified using FEA for peice of mind. A good case study.

 

[chicopee]

Will the door have stiffners or ribs?

 

[chicopee]

It seems to me that if the door was actually the ellipsoidal head and hinged on the side, then you would have something more consistent with designs I have seen in the past.

 

[saplanti]

If you are going with this design I would suggest to use spherical (will require the minimum thickness) head for easy manufacture if it does not extend the chamber length enourmously, and increase the corner radius to the largest possible on the opening if acceptable to reduce stress concentration at corner connections .

This (or your) design definetely will end up with higher wall thickness on the head and opening nozzle due to the large opening size, and can only be calculated by FEA unless you do not have previous model study or experience.

So I would expect you do your calculation by FEA, because this size of openings are not supported by the code rules.

Verifier is probably either go through your analysis or do independant FEA study using your drawings.

U-stamping or not does not reduce your responsibility in the stress analysis, and do not you thing this is an adventage in the design issue.

Kind regars,

Ibrahim Demir

 

[Crompto29]

Some good points saplanti,

The required head thickness for a large circular manway in it is 16mm. Due to the Head 'actually' having a rectangular doorway that will cause stress concentrations, I have increased the head thickness to a guestimate of 25mm.

According to the code there is no limit to the size of an opening in a head. The code then says I 'may' also use another specified method.

As I see it I have designed the opening correctly for membrane stress. But simply guestimating an extra thick head to take care of the stress intensities is not good enough.

So your right, simply adding a few mm is not enough. Further FEA analysis is required, which the 3rd party would do if I don't.

I have since carried out an FEA Analysis and using ANSYS with Inventer. It returns stress intensities in the head around the corner of the doorway that are 80% of allowable. So a good guestimate. I will include the FEA Report as an Appendix.

 

[racookpe1978]

Drop the idea of a "standard" hospital bed and use a narrower stretcher-type flat (and thinner!) mattress-on-a-platform.

Then , suspend the platform (30-36 inches wide) from a set of rollers on both sides so the patient isn't on a bed on floor-level rollers that must itself be rolled into the tank, but is on a shelf on rails.

Yes - they will have to move the patient from bed to platform.