Welded Aluminium Structure

[drawoh]

I am looking into the possibility of a welded, sheet metal aluminium chassis to be mounted in an aircraft, possibly outside. Probably, the walls will be around 1/8" (3mm) thick. Probably, I would use 5052-H32 or 5086-H32, which is bendable, and has reasonable annealed strength. This would be an optical device so it will be designed for rigidity. It will be very, very strong.

How difficult will it be to get this certified safe for flight? Has anyone any experience with this?

Another option is a casting, which I understand can be another can of worms.

I have not started modelling yet. I am staring at an arrangement in 3D[ ]CAD.

--
JHG

 

[berkshire]

In his original post Drawoh said this was a chassis. To me, that word, means a base for other things to be mounted upon. You then have the considerations of how much load do those things generate ?
B.E.

You are judged not by what you know, but by what you can do.

 

[tbuelna]

"...a welded, sheet metal aluminium chassis to be mounted in an aircraft, possibly outside."

"...this is an optical device with a laser. By the time I make it rigid enough to stay aligned under all the shock and vibration the helicopter imposes, it will be very strong..."

"I need to keep weight down. If this were a tubular space frame, there would be no advantage for aluminium over 4130 steel. This is an enclosed chassis. The wall of an aluminium chassis will be thicker and stiffer. "

"The aluminium structure would be a sealed sheet metal box. The walls would be at least .10" (2.5mm) to control warping, and to assure that there is room to grind off edges."

Based on the information provided so far, all we know is that the structure is a platform for mounting "an optical device with a laser", somewhere inside or outside on a helicopter, where weight and stiffness of the optical platform are primary concerns.

 

[drawoh]

...

"The aluminium structure would be a sealed sheet metal box. ...

Thanks.

Design has barely started, and I am looking at all my options. I have worked up a cool looking casting with a fibreglass cover. If management is determined for this to look cool, this is the way. I have contacts who rapid prototype wax patterns. Machined from billet is a good way to do one piece. It will cost more if we do a lot of them. I have designed welded sheet metal chassis in the past, but they were not mounted in aircraft. Whatever I do, I will make it optically rigid, with the optical alignment not affected by mounting in the aircraft.

Sheet metal is a near net shape fabrication method with fairly thin walls. There are ways to make it look good. I just want to know if anyone has experience handing something like this to a DAR.

Making a casting look good, making machining from billet look good and making sheet metal look good are three very different tasks. I need to pick one and stick with it.

--
JHG

 

[rb1957]

why wouldn't you 3D print the prototype ? odd to choose casting when weight was previously mentioned as a design goal.

once you define your design you may be able to see how sheet-like it is, and maybe welding or conventional assembly will be options.

welding is a perfectly valid assembly method, as long as you're aware of the pitfalls.

You've mentioned pressure ... what stress level do you expect the box to work at ?

If the structure is going to have a fiberglass fairing cover, then it can look as ugly as a (insert desired reference here) if you want. I'd make the structure efficient as I can afford, and let the fiberglass fairing look pretty.

another day in paradise, or is paradise one day closer ?

 

[drawoh]

rb1957,

I did not mention pressure. Almost certainly, this thing will have a breather desiccator to ensure there is no pressure. All I need to do is keep optics clean.

3D printing a prototype is a good idea if I plan to do a casting. I need to solve whatever problems my fabrication method causes.

The chassis is visible so far, so it need to look good. Mounting a five sided cover over a flat optical platform is a reliable way to re-align optics upon assembly, and probably upon installation as well.

--
JHG

 

[Sparweb]

For "prettiness" points, take a look at Leica airborne optical scanners. Though designed for internal mounting, the style is functional too.

STF

 

[tbuelna]

So the optical equipment will be located inside a sealed volume?

Desiccant breather systems are a great way to eliminate moisture within an enclosed volume while equalizing pressure with the outside atmosphere as it changes. Just remember that the desiccant element must be replaced on a regular basis. And the air within the sealed volume at assembly of the system will have some humidity, and should be purged using something like dry nitrogen.

 

[drawoh]

tbuelna,

We purge these things with nitrogen. Equalizing pressure is the important thing. I was shocked when I worked out the force of a sealed enclosure even at 10,000ft.

--
JHG

 

[rb1957]

"The aluminium structure would be a sealed sheet metal box." and "this thing will have a breather desiccator to ensure there is no pressure." ... it'll be a tricky control to ensure no pressure in a sealed box ... as the plane climbs you'll need to remove "air" for the box, as it descends you'll have to add it ... tricky, but not impossible (nor ludicrously complex).

If you need to visualise the structure I'd 3D print it, whether casting or not. These days I'd've thought that our 3D models were good enough representation.

another day in paradise, or is paradise one day closer ?

 

[drawoh]

rb1957,

One of our manufacturing guys told me that once I had designed something, it was his job to figure out how to manufacture it, and he could do it any way he damn well pleased. Strictly speaking, he was right, and almost certainly, he knew better.

In over thirty years of mechanical design, I do not recall ever designing something significant where I did not know how it would be fabricated. Usually, this would be machining. Obviously, it must be feasible to fabricate your part by your chosen method. Less obviously, you want to take advantage of your manufacturing technology. Machined parts can be very accurate. Sheet metal allows you to design thin-walled, 3D, near net shapes at low cost. Castings can contain all sorts of weird curves, and can be very complex with minimal cost impact. In all cases, you have to be able to get tooling in and out, including your welder. In many cases, you have an up-front tooling or programming cost.

Any time I have a structure or mount at the point where it can be visualized, I know how it will be fabricated. Right now, I am visualizing a casting, with an FRP[ ]cover. I have started off fabrication drawings, I have done some tolerance stacks, and it looks like I am in trouble. I need to figure out accurate interface surfaces, as well as airworthiness. Early in the design process is the time to have these problems. I can adapt the design to solve them, or at worst, I can switch to another design concept.

--
JHG

 

[rb1957]

there is a cart and a horse ...

designing something without an eye to how it's going to manufactured is a short cut to a redesign.

your ME/planner is a little off base ... if you design it with welds, it has to be welded. he can come back to you and bitch and demand a design change. we'll often get design change requests "we want to make this a machining, rather than s/m, 'cause the s/m shop is full and the machine shop needs work" ... sign

we get the same comments here ... talk to one guy in the s/m shop about how to do something, then the next guy says "I can't bend this" ... "well, Benny said he could" ... "you can't design a structure for a specific mech to build for you" ... sigh (go back and reread every Dilbert, again).



another day in paradise, or is paradise one day closer ?

 

[PBPouw]

Why not a simple sheet metal riveted enclosure ?
- cheap
- easy to do stress calculations
- easy to modify if necessary
- very simple to add local re-enforcements, stringer, angles etc....

I mean we have been building rivet sealed airframes for more then 80 years !

 

[IRstuff]

You don't say anything about alignment accuracy and retention. Requirements for these in optical systems tend to drive to 6061. We tend to CNC stuff, because it's presumably easier to anneal out all the stresses so that alignment doesn't shift over time/temperature. For our systems, flatness and planarity are absolute requirements, so even if you can do a bang-up job of welding/annealing, it's likely that you'll still have to lap mating interfaces to achieve and maintain the desired initial alignments. We've had systems that required lapping because machining stresses couldn't be completely removed.

Note that everything is relative, including stiffness. While Al has great stiffness for what it's supposed to do, it's not really stiff enough for some applications, where you need to go something like steel.

TTFN
I can do absolutely anything. I'm an expert!
homework forum: //

http://www.engineering.com/AskForum/aff/32.aspx

faq731-376 forum1529