What's new
What's new

How do I calculate the correct width of skin removal when bending an aluminum honeycomb panel?

judgedredd20

Plastic
Joined
Oct 15, 2023
Location
Niagara Region, Ontario, Canada
This is the first time I am posting here so I'm not sure if this is the correct sub-forum to post to.

I have an aluminum honeycomb panel with these dimensions:
skin 1 - 0.8mm thick
honeycomb material - aluminum small cell (1/4") 8mm tall
skin 2 - 1mm thick

I need to bend this panel 90 degrees with the 1mm skin on the inside of the bend. What is the width of the channel (only going down 1mm to remove the skin, not the honeycomb) that I have to remove of the 1mm skin to be able to do this? I have seen a YouTube channel here:
where the bending technique is shown. What is not shown is how to calculate how much aluminum skin to remove for different angled bends. In the video both skins are 0.7mm and the honeycomb height is 10mm. The width of the removed area is 26.4mm for the 90 degree bend, which seems counter-intuitive. I thought it would be the radius of the circle of the outer skin divided by 4 (the radius being the height of the honeycomb, which is 10mm in this case).

Does anyone know how to go about doing this calculation for my piece?
 
Last edited:
This is the first time I am posting here so I'm not sure if this is the correct sub-forum to post to.

I have an aluminum honeycomb panel with these dimensions:
skin 1 - 0.8mm thick
honeycomb material - aluminum small cell (1/4") 8mm tall
skin 2 - 1mm thick

I need to bend this panel 90 degrees with the 1mm skin on the inside of the bend. What is the width of the channel (only going down 1mm to remove the skin, not the honeycomb) that I have to remove of the 1mm skin to be able to do this? I have seen a YouTube channel here:
where the bending technique is shown. What is not shown is how to calculate how much aluminum skin to remove for different angled bends. In the video both skins are 0.7mm and the honeycomb height is 10mm. The width of the removed area is 26.4mm for the 90 degree bend, which seems counter-intuitive. I thought it would be the radius of the circle of the outer skin divided by 4.

Does anyone know how to go about doing this calculation for my piece?
Not critical as you will need to glue and rivet a doubler on the inside corner anyway.
 
The answer is probably 'none'. If the skin is only 1MM or less, it's gonna bend just fine. My bigger concern would be how do you prevent the honeycomb from collapsing under the bend pressure. It's a lot taller and weaker than than the solid skin.

Plus, if you were to reduce the skin thickness for easier bending, you'll substantially reduce the strength of the final product...the skin is what's keeping the honeycomb in position.
 
Thanks for both of you responding.
moonlight machine, I will be using 1" right angled aluminum extrusions as reinforcements on the inside of the bend and will use structural glue only and no rivets (as its not going to be a chassis component, or any other structural component). The glue will have a shear strength of over 4,000 lbs/psi. I would like to get the width of the material to remove precise so that the surface area to be glued is as large as possible. Plus its the feeling of doing the job right. It'll bother me if there are gaps between the inside surfaces, even though it will not be seen.

GregSY, what you suggest won't be possible. Have you ever bent corrugated cardboard? The inside bend always crinkles and warps the inside layer of cardboard there. That's why the layer needs to be cut out and reinforced afterwards.
 
The answer is probably 'none'. If the skin is only 1MM or less, it's gonna bend just fine. My bigger concern would be how do you prevent the honeycomb from collapsing under the bend pressure. It's a lot taller and weaker than than the solid skin.

Plus, if you were to reduce the skin thickness for easier bending, you'll substantially reduce the strength of the final product...the skin is what's keeping the honeycomb in position.
You cut the inside skin and remove the honeycomb at the bend. Once the skin and honeycomb are removed you can usually just fold it by hand as it is so stiff on each side of the cut. I used to do a fair bit of aluminum honeycomb work back in the day, before carbon fibre took over Indy Cars.
 
You cut the inside skin and remove the honeycomb at the bend. Once the skin and honeycomb are removed you can usually just fold it by hand as it is so stiff on each side of the cut. I used to do a fair bit of aluminum honeycomb work back in the day, before carbon fibre took over Indy Cars.
Thanks for the reply. Did you do any calculations for the amount of skin to remove, in terms of width, when you were doing this? If so, how did you do the calculations? Its the formula that I am looking for.
 
Thanks for the reply. Did you do any calculations for the amount of skin to remove, in terms of width, when you were doing this? If so, how did you do the calculations? Its the formula that I am looking for.
Trig is likely to have the answer for ya, providing you want a tight bend. I would start with how wide you need to make it to get a 90 degree bend, and adjust as required to get different angles.
 
Seems like we used to make the slot about the same width as the thickness of the honeycomb. We used a router against a straightedge to do the slot. We used redux 410 adhesive. It's probably obsolete now.
 
moonlight machine, that's what I thought but when looking at the video the slot he made for a 90 degree bend was 26.4mm for a honeycomb thickness of 10mm. This is what confuses me. So I thought that there must be some formula to do these calculations. Mind you, he left the honeycomb in place, just removing the skin layer which was 0.7mm. By leaving the honeycomb in place it created a larger radius on the outside of the bend. This is what I would like, not a sharp bend.

Thanks trevj for responding.
 
Via prof YouTube examples the better bends follow a pattern.
The skim width is the length of arc on outside radius-
Where outside radius is calculated to be total thickness + 1 or 2 skin widths on 90
Calculate at + 0 or 1 skin width at 45.

I do not see a common neutral axis or calculated material axis in this- but honeycomb is very unknown and non-constant cross section and it is aluminum which is a metal with cooties.


Trig lovers: calculated radius is total thickness + 2sin(alpha)x skin thickness.
 
Last edited:
memphisjed, thanks for responding. I went back to the video I posted and the widths of the skim is 10.9mm for the 45 degree bend, which just about lines up with your last calculation, and the 90 degree bend had a skim width of 26.4mm. This is very different from what you posted. When watching the video and seeing the 90 degree bend done it leaves me very puzzled as to the difference in the skim width with the 45 degree bend, yet it has the desired result.
 
Last edited:
.548 proper units + 2*.07 imaginary units = .604 ish. Arc length of .604 radius with 90 included = .946 proper units. That’s 24 mm. Not far off for initial calculations.

I have compensated length adjustments for different styles of bending found by error (lots of errors) but none are drastically different from calculated proportions. You find those values over time.
 
Shouldn't that be .688 ish?

So for my panel, which has 1mm skin, 8mm of honeycomb height and 0.8mm of skin, the calculation would be:
.548 + 0.1+ 0.08 = 0.728... Arc length of .728 radius - the rest I'm not able to compute ...

I take it you're converting .946 from inches to mm?
 
Last edited:
.548 inches plus (.1 + .08 )mm = .555 inches

Your total thickness is 9.5 mm.
Radius = 9.5 + 1.8 mm => arc length of 17.75 mm on 90 degree arc.

It is important to measure your cuts and skims on test parts. Measure what you skimmed, not what you tried to do. Also measure legs after bends. These leg results tell you where to start bends (leg a) and how long to cut parts or start next bends (leg b). The cut width you can then note and tune to next bend.

Bending metal is easy, bending to measurements repeatabley is not so easy.
 
Ok. There seems to be some confusion here...
In your original calculation you gave:
.548 + 2*.07
I took the .07 to be the 0.7mm skin in the video example, so I followed that format.

The total thickness of my panel is:
1mm (skin 1 thickness) +
8mm (honeycomb height) +
0.8mm (skin 2 thickness)
= 9.8mm

So new radius is 9.8 + 1.8 => arc length ?? (I have no idea how to calculate that!)
 
.548 inches plus (.1 + .08 )mm = .555 inches

Your total thickness is 9.5 mm.
Radius = 9.5 + 1.8 mm => arc length of 17.75 mm on 90 degree arc.

It is important to measure your cuts and skims on test parts. Measure what you skimmed, not what you tried to do. Also measure legs after bends. These leg results tell you where to start bends (leg a) and how long to cut parts or start next bends (leg b). The cut width you can then note and tune to next bend.

Bending metal is easy, bending to measurements repeatabley is not so easy.
In aircraft sheet metalwork, we dealt with "K-Factor", either looked up on charts, or calculated out the long way, for accounting for the amount of material that was required to make a bend of a specific radius, on a specific thickness of material. I would suggest looking in to that, as a starting spot, then experimenting on some scrap. Where you are referencing off of matters, unless you can do all your work without having to favor one end or the other, and still get your dimensions. May or may not apply to your situation, but you had to plan your bend sequence pretty carfully, if you were making complex nested sheet parts!

My dealings with Honeycomb were generally of the Repair a hole type work, or document where some idiot kicked the mud off his boots as he climbed up the steps to the rotor head... <spit>

In any case, it might be something worth looking in to, if not producing a correct result, it should get you closer.

IIRC, we used some foaming adhesive to infill the gaps, when we were fitting panels together in school. Used heat to set the stuff up. No idea what it was... Been a few years...
 
Hi trevj,
Thanks for responding. I am cutting out the 1mm skin and having that on the inside as the 0.7mm skin is easier to bend! On the other hand having the 1mm skin on the outside will make the item more durable. I am assuming that the width of cut to remove the skin will depend on which side of the honeycomb is on the outside of the bend.
 
In aircraft sheet metalwork, we dealt with "K-Factor", either looked up on charts, or calculated out the long way, for accounting for the amount of material that was required to make a bend of a specific radius, on a specific thickness of material. I would suggest looking in to that, as a starting spot, then experimenting on some scrap. Where you are referencing off of matters, unless you can do all your work without having to favor one end or the other, and still get your dimensions. May or may not apply to your situation, but you had to plan your bend sequence pretty carfully, if you were making complex nested sheet parts!

My dealings with Honeycomb were generally of the Repair a hole type work, or document where some idiot kicked the mud off his boots as he climbed up the steps to the rotor head... <spit>

In any case, it might be something worth looking in to, if not producing a correct result, it should get you closer.

IIRC, we used some foaming adhesive to infill the gaps, when we were fitting panels together in school. Used heat to set the stuff up. No idea what it was... Been a few years...
K factor I use on pipe/hollow structures. I figure length between bends off neutral axis, length of material needed for bend from k-axis. Solids share k/neutral. This honeycomb with skimming is one we’re I would need lots more sample size and pre bend/post bend measurements to guesstimate k axis.

Length of arc = 2*pi*r x Angle/360 , or right click arc in cad and it tells you length.
 
Ahha! So my arc length would be:
(2*pi*(9.8+1.8) x 90)/360 = 18.22mm = width of channel.

Applying this to the video:
11.4 + 1.4 = 12.8
(2*pi*12.8)x90/360 = 20.1mm

yet they have 26.4mm.

I have emailed the head of the dept of Swansea Uni to find out how they calculated these cuts.
 
Odd, Swansea university is the first one I went to, they have a new campus since my visit ( very nice it is) we had the old “school of mines” , an ancient building.
New stuff there was that thrust ssc car designs and stuff.
Don’t know many there now, however they are fairly helpful imho.
Mark
 








 
Back
Top