Door design

[galderdi]

As many of you know I am a beginner undertaking a fairly ambitious build of a 550 Spyder. It is not a true replica. So long as the average car enthusiast can see what it is I will be happy. So I can take a few liberties. One of the areas I am trying to work out is the door design.
Both the original and my car will have rather shallow doors with large/deep sills. The skin of the door needs to come a long way over the top of the door and will sit very close to the occupant's outside shoulder. For this reason I want to avoid having a sharp 90 edge to thin inside of the door skin.
I want to curve the skin under itself based on a 1inch pipe (1/2 inch radius).
There will be a fairly traditional door frame and the skin will attach in the usual way. That is tipping the edge of the skin all the way around and clamping the skin over the edge of the door frame.
But that method requires access to the edge to hammer the skin down. That will work fine all the way around until I get to the 1 inch curve and it turns back on itself.
I have a few ideas but non of them are ideal.
I am fairly sure (unless someone has a better idea) The edge of the skin will proceed as normal up to the top but needs to be out of the way once it gets to the horizontal plane on top of the door. That way I can still get the hammer to the inside of the edge to secure the skin. The attached sketch is not to scale. But the area I am talking about is where the question mark is pointing. If the edge of the skin wraps around past the dark mark then it will be in the way when trying to secure the top of the skin.

The best I have come up with so far is to leave a piece missing from the corner of the skin at each end (A in the diagram). This will allow me to secure the main part of the skin (C in the diagram). Then I will have a secondary piece of the door frame which will have a vertical piece at each end so it can be bolted or riveted to the lower part of the frame (B in the diagram), and a horizontal piece with the same 1 inch diameter running the length of the car which can slide under the skin to give the structure support. Then I would need to decide whether the skin should weld to it or some other method. I am assuming the same approach would be applied to the front and rear corners of each door with this support piece running the full length of the door joining the tops of the two end frames.

Any better ideas?
Any examples of a similar project?
Any pit falls I might have overlooked?

[Moving molecules .]

Like most early cars the frame is steel and the Door skin is Aluminium, are you using all Aluminium.

[Moving molecules .]

[cliffrod]

When I have to produce a difficult aspect of a shape, one that looks obvious but is somehow technically difficult or confusing, I stop and compare it to other integral elements. Because they are integral, it may be possible to change the order of operations to produce the hardest or most problematic portion earlier and leave another less difficult aspect(s) to follow. On certain jobs- especially in stone but sometimes in metal- what remains untouched becomes the finished detail as work around it is resolved. It's also good to think about what will and will not be visible on the finished job. If I have to mangle or fight a region of a project, I want it to be less visible or prominent. Sometimes these two perspectives can work together.

Is it feasible to shape the door skin and then develop this difficult & very visible portion as much as possible before installing it over the door frame while leaving other portions along sides and/or bottom to be completed once the skin is on the frame? If you do decide to cobble it up, locate that work where is is less likely to be seen versus right on top and easily seen.

The vertical to horizontal transition on the sides of this RD350 seat I'm (slowly) working on proved to be a challenge to do well as a single piece. Cleaning up what I started didn't make sense. I decided to go a different direction. Two new pieces were cut for the sides, folded with the soft radius I wanted and then each were quickly tuned into the proper mild S shape to follow the frame with the shrinker and stretcher. I'll place my beginner-quality welds under the upholstery where they will not be seen. Sides and transition are now new-metal flawless. very cool.

For door skins like yours, I would try to develop the wrapped area that's challenging you early, then install it over the frame and work on folding the rest into place around the sides & bottom. After the top portion, make the door latch area the prettiest, then the hidden hinge side and finishing on the bottom where few people will ever look.

[Moving molecules .]

Life is very strange...

[galderdi]

Like most early cars the frame is steel and the Door skin is Aluminium, are you using all Aluminium.

I am going with an all aluminium frame with the exception of the steel hinges, steel door striker, steel intrusion bar between them and steel structure to support the hinges and striker.

[galderdi]

When I have to produce a difficult aspect of a shape, one that looks obvious but is somehow technically difficult or confusing, I stop and compare it to other integral elements. Because they are integral, it may be possible to change the order of operations to produce the hardest or most problematic portion earlier and leave another less difficult aspect(s) to follow. On certain jobs- especially in stone but sometimes in metal- what remains untouched becomes the finished detail as work around it is resolved. It's also good to think about what will and will not be visible on the finished job. If I have to mangle or fight a region of a project, I want it to be less visible or prominent. Sometimes these two perspectives can work together.

Is it feasible to shape the door skin and then develop this difficult & very visible portion as much as possible before installing it over the door frame while leaving other portions along sides and/or bottom to be completed once the skin is on the frame? If you do decide to cobble it up, locate that work where is is less likely to be seen versus right on top and easily seen.

The vertical to horizontal transition on the sides of this RD350 seat I'm (slowly) working on proved to be a challenge to do well as a single piece. Cleaning up what I started didn't make sense. I decided to go a different direction. Two new pieces were cut for the sides, folded with the soft radius I wanted and then each were quickly tuned into the proper mild S shape to follow the frame with the shrinker and stretcher. I'll place my beginner-quality welds under the upholstery where they will not be seen. Sides and transition are now new-metal flawless. very cool.

For door skins like yours, I would try to develop the wrapped area that's challenging you early, then install it over the frame and work on folding the rest into place around the sides & bottom. After the top portion, make the door latch area the prettiest, then the hidden hinge side and finishing on the bottom where few people will ever look.

You are thinking along similar lines to me. Its not so much about making it pretty. It more about a design that I can actually construct at all. Ideally the design would be simple with the rolled top of the skin all the way along. But that just doesn't work as I wouldn't be able to secure the skin to the frame in that tight corner. So yes producing a secondary piece to be added after the skin is along the same lines as my thinking. But even then it would leave a join where I don't want a join. I could, and probably will, weld the join. But it seems unusual to weld a door skin and I'll need to be really careful about warping. So I Was hoping there was a better way.

[neilb]

would it be possible to rivet the skin along the top of the door, i'm thinking the skin has rolled over on towards the top of the inner frame. the rivets could be covered with a trim strip of leather to hide the rivets. not so much pop rivets but more of the flush fit self piercing rivets we use on aluminium panels on the jags an range rovers of today

[galderdi]

Yes I think rivets will play a part somewhere. I'm not familiar with the type you've mentioned but I'll check them out. If I use normal rivets they would be along the underside of the edge so it isn't readily visible (unless you lower your head below the door)

[neilb]

have a look at this video, I've not watched it, it was just the first one that came up

https://youtu.be/2Ws3BWXMWbQ

these are the manual type rivet guns, we use an air/hydraulic type unit

[neilb]

a couple of pics of what they look like and look like installed