Background
I have the diabolical drawer. I'm very happy with it in all aspects but one. I put 4 stock Grenadier tires(on steelies) on the floor and it deflected (bent).. clearly when i opened the drawer but otherwise it rested on the lip of the drawer which seemed fine.
I talked to Alan at diabolical about it and it is possible that we miscommunicated but i think he is basically not really disclosing the weight bearing capacity on the drawer when open, but he states 500lbs when closed and designed to sleep on. My math works out to about those numbers for static weight and i think that is really the design parameter.
Having said that, i don't have my load evenly distributed across the entire plate (i find that incredibly hard to do) and i also want to use it loaded when driving so i am chasing dynamic weight and high amount of weight concentrated to the center of the plate.
the calculations below all focus in on solving for that use case.
Fixing my deflection (or bend in the metal plate) problem
Being the trouble maker that i am i set myself out on a journey to figure out how to reinforce the drawer for the weights i want to place in it... Lets say 4 really heavy Grenadier tires
I have done a bunch of math on this ( and structural engineering isn't my strong suit so i had to reach textbooks).
Having said that i have implemented this solution and it works well in my car. I have loaded up my car and i can see no deflection (bend) where previous there was a significant bend.
Here is how i analyzed it, and what i did.
Structural input
The drawer comes with an 1/8" aluminum top plate. The system consists of two 40.5" x 17" plates supported by approx 1" side mounts, a 500 lb-rated front drawer ledge, and a 2" rear vertical bend. The plates have a 90 degree bend 3/8" down the center where they mate. I didn't really spend too much time measuring the side mounts since the issue is really the unsupported part of the plate (or the area in between) so if i got a measurement wrong on the side mount, i did those from memory as i typed this up. They are not the key part i analyzed, the plate itself is.
In all calculations i've included if the top plate rests on the closed drawer lip but critically, i have chosen not to distribute the weight evenly across the top plate. Note: I've chosen to assume that most of the weight is concentrated to the middle 30% of the plate front to back. Its a tougher standard but i was more worried about weights not being uniformly applied across the plate and more concentrated. And nothing will we worse than placing it all down the middle.
Structural Evolution
Option 1: Initial Design as it came out of the box
Two 35" L shaped aluminum re-inforcement bars + 90 degree bend in front and rear. Rear most plate has a 2" bend.
- Structure: 1/8" Aluminum Plate reinforced with two 35" L-bars (3/8" tall) via 5 welded 6mm threaded studs per bar (I'm not sure if they are 6mm, i didn't measure but it doesn't matter so much as long as they are not tiny like 3mm screws and they are not)
Note this is with the drawer closed and assuming the front lip is resting on the drawer.
Option 2: Steel Spine (35" Length)
- Structure: 35" A36 Steel U-Channel (2-1/2" x 5/8" x 3/16") bonded with MA310 structural adhesive at the center of each plate. Flat surface of U channel bonded to plate. Length identical to the L beams.
Option 3: Edge-to-Edge Reinforcement eg Full span steel (Final Build)
- Structure: Steel U-channel extended all the way to each side mount ledge, eliminating any gap across the panel. Since the steel can't sit on the edge mount, the 1/8" aluminum plate acts as the "bridge" for that last bit of distance, that section is directly supported by the mounting rail underneath, the aluminum isn't "bridging" a gap in the air anymore. It is being squeezed between the steel spine and the mounting rail. This type of of weight bearing is called compression, and aluminum is much stronger in compression than it is in bending.
This is 20x stiffer than stock
It is 8x stronger in terms of weight capacity assuming the side mounts can handle the load.
It adds 20lbs of weight
However... you loose static load if you open the drawer and only make it supported on 3 sides.
if you open the drawer its now only supported on 3 sides. the static load it can handle drops to 400lbs
.
if you want to keep the static load high you need to add one more steel beam to the be right behind the front lip of the top plate. Same dimensions as the other beam.
if you do this and glue it in the double steel beam on the front plate will let you put 850 lbs of weigth on top of the lid with the drawer open. Its what i am considering to do next to really finish off my project. So in the end, you may want to glue in 3 re-inforcement channels.
I would be remiss if i didn't say that i spent weeks calculating the ideal material and shape to get to this. Mostly because i am unfamiliar with the subject, but also shapes matters greatly and practical experiences would have steered me away from some bad choices i considered.
Technical Analysis
I picked MA310 for its vibration properties and its strength + that it works well for surfaces that aren't perfectly even.
lets say i would have gone with 6x M10 bolts. Grade 8.8 metric bolts has a steel shear strength of 60,000 PSI vs 3,000 PSI for the glue. But the issue is point strength vs full surface strengths. Those bolts has a combined shear capacity of 27,000PS vs the shear capacity of the glue at 262,500 PSI (assuming you prepped the surface correctly). I would take the structural glue any day of the week, especially for this location.
Steel/Adhesive Bond
The MA310 Structural Adhesive allows the assembly to function as a composite beam which is the trick that makes thia work. It's the same trick used in designing modern aircraft wings with composites. The adhesive distributes the load across 87.5 square inches, preventing point-stress failures that happens in bolted systems. The other nice effect with MA310 is that It is also a vibration-damped bond thanks to the methacrylate adhesive.
For MA310 to bite i used a zirconia flap disc on my angle grinder and scuffed it up on both sides. I was careful to prime and paint the steel since i don't like the idea of dielectric materials (which realistically is only a risk if someone spills salt water on it but i did it anyway). The glue isolates the materials from each other.
Gap Factor
In Option 2, the gap between the steel channel and the plates mounts acts as a lever arm. Under a 400 lb load at 3G impact, stress reaches ~37,000 psi, which is the aluminum's yield point. Option 3 eliminates this risk by extending the steel to the mounts. This can be added afterwards, just glue the steel close to the already in place beam if you don't have enough.
Installing a full width beam makes the whole system much, much stronger.
A few more thoughts
With the plates bolted together and the steel extended, this plate now exceeds the strength of many (most?) commercial overland drawer tops. The primary constraint is now the mounting hardware on the sides rather than the floor itself but if that happens i think thats a trivial place to add stiffening to.
Shopping list:
Pics or it didn't happen...
It was raining and cold, i managed to get the paint on the day before when it was warmer and dry but when time to glue home i hijacked the kitchen island.
I have the diabolical drawer. I'm very happy with it in all aspects but one. I put 4 stock Grenadier tires(on steelies) on the floor and it deflected (bent).. clearly when i opened the drawer but otherwise it rested on the lip of the drawer which seemed fine.
I talked to Alan at diabolical about it and it is possible that we miscommunicated but i think he is basically not really disclosing the weight bearing capacity on the drawer when open, but he states 500lbs when closed and designed to sleep on. My math works out to about those numbers for static weight and i think that is really the design parameter.
Having said that, i don't have my load evenly distributed across the entire plate (i find that incredibly hard to do) and i also want to use it loaded when driving so i am chasing dynamic weight and high amount of weight concentrated to the center of the plate.
the calculations below all focus in on solving for that use case.
Fixing my deflection (or bend in the metal plate) problem
Being the trouble maker that i am i set myself out on a journey to figure out how to reinforce the drawer for the weights i want to place in it... Lets say 4 really heavy Grenadier tires
I have done a bunch of math on this ( and structural engineering isn't my strong suit so i had to reach textbooks).
Having said that i have implemented this solution and it works well in my car. I have loaded up my car and i can see no deflection (bend) where previous there was a significant bend.
Here is how i analyzed it, and what i did.
Structural input
The drawer comes with an 1/8" aluminum top plate. The system consists of two 40.5" x 17" plates supported by approx 1" side mounts, a 500 lb-rated front drawer ledge, and a 2" rear vertical bend. The plates have a 90 degree bend 3/8" down the center where they mate. I didn't really spend too much time measuring the side mounts since the issue is really the unsupported part of the plate (or the area in between) so if i got a measurement wrong on the side mount, i did those from memory as i typed this up. They are not the key part i analyzed, the plate itself is.
In all calculations i've included if the top plate rests on the closed drawer lip but critically, i have chosen not to distribute the weight evenly across the top plate. Note: I've chosen to assume that most of the weight is concentrated to the middle 30% of the plate front to back. Its a tougher standard but i was more worried about weights not being uniformly applied across the plate and more concentrated. And nothing will we worse than placing it all down the middle.
Structural Evolution
Option 1: Initial Design as it came out of the box
Two 35" L shaped aluminum re-inforcement bars + 90 degree bend in front and rear. Rear most plate has a 2" bend.
- Structure: 1/8" Aluminum Plate reinforced with two 35" L-bars (3/8" tall) via 5 welded 6mm threaded studs per bar (I'm not sure if they are 6mm, i didn't measure but it doesn't matter so much as long as they are not tiny like 3mm screws and they are not)
| Scenario | Static load capacity if placed in the middle 30% of the plate | Dynamic load capacity in the middle 30% of the plate | Result |
|---|---|---|---|
| Plates not connected | 220 lbs 350lbs if evenly distributed | 75 lbs 125lbs if evenly distributed | bowing and friction at the seam |
| Plates bolted at seam | 280lbs 450 lbs if evenly distributed | 110 lbs 175 lbs if evenly distributed | reduced sag, significant flex |
Option 2: Steel Spine (35" Length)
- Structure: 35" A36 Steel U-Channel (2-1/2" x 5/8" x 3/16") bonded with MA310 structural adhesive at the center of each plate. Flat surface of U channel bonded to plate. Length identical to the L beams.
| Scenario | Static load capacity in the middle 30% of the plate | Dynamic load capacity in the middle 30% of the plate | Result |
|---|---|---|---|
| Plates not connected | 650lbs 900 lbs if evenly distributed | 225 lbs 325 lbs if evenly distributed | Rigid center; unreinforced side gaps create "hinge" risk |
| Plates bolted at seam | 850 lbs 1,200 lbs if evenly distributed | 350 lbs 450 lbs if evenly distributed | High rigidity (*my current state) under 3mm deflection at 400 lbs placed centered 30% across the plate from front to back. |
Option 3: Edge-to-Edge Reinforcement eg Full span steel (Final Build)
- Structure: Steel U-channel extended all the way to each side mount ledge, eliminating any gap across the panel. Since the steel can't sit on the edge mount, the 1/8" aluminum plate acts as the "bridge" for that last bit of distance, that section is directly supported by the mounting rail underneath, the aluminum isn't "bridging" a gap in the air anymore. It is being squeezed between the steel spine and the mounting rail. This type of of weight bearing is called compression, and aluminum is much stronger in compression than it is in bending.
| Scenario | Static load capacity in the middle 30% of the plate | Dynamic load capacity in the middle 30% of the plate | Result |
|---|---|---|---|
| Plates not connected | 900 lbs 1,400 lbs if evenly distributed | 400 lbs 550 lbs if evenly distributed | Weight transfers directly to mounts; no hinge effec |
| Plates bolted at seam | 1,200 lbs 1,800+ lbs if evenly distributed | 500+ lbs 750+ lbs if evenly distributed | Stiffness increased by ~1,700% over stock |
This is 20x stiffer than stock
It is 8x stronger in terms of weight capacity assuming the side mounts can handle the load.
It adds 20lbs of weight
However... you loose static load if you open the drawer and only make it supported on 3 sides.
if you open the drawer its now only supported on 3 sides. the static load it can handle drops to 400lbs
.
if you want to keep the static load high you need to add one more steel beam to the be right behind the front lip of the top plate. Same dimensions as the other beam.
if you do this and glue it in the double steel beam on the front plate will let you put 850 lbs of weigth on top of the lid with the drawer open. Its what i am considering to do next to really finish off my project. So in the end, you may want to glue in 3 re-inforcement channels.
I would be remiss if i didn't say that i spent weeks calculating the ideal material and shape to get to this. Mostly because i am unfamiliar with the subject, but also shapes matters greatly and practical experiences would have steered me away from some bad choices i considered.
Technical Analysis
I picked MA310 for its vibration properties and its strength + that it works well for surfaces that aren't perfectly even.
lets say i would have gone with 6x M10 bolts. Grade 8.8 metric bolts has a steel shear strength of 60,000 PSI vs 3,000 PSI for the glue. But the issue is point strength vs full surface strengths. Those bolts has a combined shear capacity of 27,000PS vs the shear capacity of the glue at 262,500 PSI (assuming you prepped the surface correctly). I would take the structural glue any day of the week, especially for this location.
Steel/Adhesive Bond
The MA310 Structural Adhesive allows the assembly to function as a composite beam which is the trick that makes thia work. It's the same trick used in designing modern aircraft wings with composites. The adhesive distributes the load across 87.5 square inches, preventing point-stress failures that happens in bolted systems. The other nice effect with MA310 is that It is also a vibration-damped bond thanks to the methacrylate adhesive.
For MA310 to bite i used a zirconia flap disc on my angle grinder and scuffed it up on both sides. I was careful to prime and paint the steel since i don't like the idea of dielectric materials (which realistically is only a risk if someone spills salt water on it but i did it anyway). The glue isolates the materials from each other.
Gap Factor
In Option 2, the gap between the steel channel and the plates mounts acts as a lever arm. Under a 400 lb load at 3G impact, stress reaches ~37,000 psi, which is the aluminum's yield point. Option 3 eliminates this risk by extending the steel to the mounts. This can be added afterwards, just glue the steel close to the already in place beam if you don't have enough.
Installing a full width beam makes the whole system much, much stronger.
A few more thoughts
With the plates bolted together and the steel extended, this plate now exceeds the strength of many (most?) commercial overland drawer tops. The primary constraint is now the mounting hardware on the sides rather than the floor itself but if that happens i think thats a trivial place to add stiffening to.
Shopping list:
- 2 x 4ft 2-1/2" x 5/8" x 3/16" A36 Steel Bar Channel (i got mine from metaldepot.com )
- 50ml MA310 (i got it from amazon.com )
- 1 zirconia flap wheel (Amazon)
- A rattle can of primer, a rattle can of flat black (my local ace hardware)
- Clamps, and a 2x4 piece of wood for use with c clamps to put some pressure on the U channel as it is glued to the plate.
Pics or it didn't happen...
It was raining and cold, i managed to get the paint on the day before when it was warmer and dry but when time to glue home i hijacked the kitchen island.
Last edited:
