What's new
What's new

OT? thermal expansion of stainless steel glued to mdf

dian

Titanium
Joined
Feb 22, 2010
Location
ch
last summer when it was 20-25°c i made a work bench for outside use. i epoxied a 0.5mm ss sheet on top of a mdf board. the dimensions are 120x60 cm.

now during the day (around 20°c) the board arches upwards by 2-3 cm. while the steel sheet gets warmer than that from the sun i still cant understand whats happening. during the night its flat. thermal coefficients seem to be: mdf 12, ss 16, epoxy 55. these are µ/m,k right? so would 4µ times 20-30°k times 1.2m = 100-150µ cause this arching? i assume the epoxy layer has nothing to do with it because of its low modulus. it should mitigate the effect.

i was planning to bolt the board down to the steel frame, but now i cant.
 
I'd guess moisture absorption and expansion of the MDF is what's dominating here. If you can allow the assembly to dry thoroughly, then try to "seal" the MDF with a liquid epoxy coating all over perhaps that may stabilize it.

Alternatively, just allow to flatten, then bond a second sheet of stainless like the first to the bottom. Now it should be pretty stable regardless of temperature/humidity, although excess moisture might degrade the MDF at some point. I'd still seal the exposed edges.
 
last summer when it was 20-25°c i made a work bench for outside use. i epoxied a 0.5mm ss sheet on top of a mdf board. the dimensions are 120x60 cm.

now during the day (around 20°c) the board arches upwards by 2-3 cm. while the steel sheet gets warmer than that from the sun i still cant understand whats happening. during the night its flat. thermal coefficients seem to be: mdf 12, ss 16, epoxy 55. these are µ/m,k right? so would 4µ times 20-30°k times 1.2m = 100-150µ cause this arching? i assume the epoxy layer has nothing to do with it because of its low modulus. it should mitigate the effect.

i was planning to bolt the board down to the steel frame, but now i cant.

The only part I "can't understand" is why you expected any other outcome?

Bit large for a classical "bi-metallic" thermostat element. But at least you only used METAL for the one element!

You want a "better way"?

- first, mdf is not your friend. Laminated Oak, Maple, or Bamboo could be, If good-enough, they will need seriously long and good drying .. then sealing,

- second you do NOT "bond" the sheet steel to the sound and shock deadener wood/grass at all.

- you place a thin resilient slip layer it will not, and CAN NOT bond to. Common polyethylene plastic laminate floor vapour barrier/underlay goods work for me. "Feltbeater" poly-olefin roofing roll goods should do fine as well.

NO ADHESIVES need apply! It WILL ALL move with temp delta. It must ALL be permitted to do so unimpeded.

"Wrap-under" at each edge with space to move, all four sides, and yer "floating" top is now independent of stressing the damping layer. And the reverse.

Next-up? Put outdoors? You need at last a shaded area or a "fly tent" to keep the direct SUN off it. I have an ordinary carport. The ambient AIR temp swings are not as challenging. And they are slower.

Differential thermal coefficient's of expansion? Fight those?

Pilgrim? The Appalachians were once higher than the Berner Oberland or even the Hindu Kush are today.

Waddyah think an incredibly PATIENT and enduring PERSISTENT "Mother Nature" utilized to LEVEL those once proud sumbiches? You figure the Shahara, the Gobi, or any other sandy or volcanic rock particle desert began life that way? Or is the PRODUCT of differential thermal erosion?

And you want to arm-wrestle with the owd b**ch over thermals of an ignorant BENCHTOP?

She'll have your ass! first time, next time, every time, ALL the time, several billions of years... and counting!

:)
 
Another vote for moisture.

Wouldn't rule it out, but WTF?

I'm 30 years with the same right-decent closed-cell foam-filled steel front door as faces almost due South. And has always needed a bit of Black Magic to adjust the deadbolt latch because the sumbich bows throughout the seasons of the year and the hours of the day from differential expansion just between indoor skin and outdoor one.

MDF, even "bone-dry forever" don't match Stainless Steel. Gots to be able to slip, somehow.

And/or be kept in dead-steady temp and humidity. Shop? Maybe. "Outdoors? Hardly!

WTH, even two disparate Inox alloys can be selected to differ by enough to do USEFUL things with. See the multiple-bar countervailing alloy shaft on a high-precision Mercury-Pendulum French-pattern mantle clock of days gone by.

Yah don't FIGHT Mother Nature over the Laws of Physics.

You put her dead-reliable ass on the payroll, and whore it out for advantage, is whatcha DO. She's good with that. Get's bored and cranky, she ain't put to work.

And then we gets floods and Earthquakes.

:D
 
I don't think the moisture effect would follow day/night cycle, its too quick

The error in the design seems the use of epoxy to glue the SS to the board, if I were to glue it at all, I'd use much more flexible glue, 2 part polyurethane used for wood floor boards glued to heated concrete floors for example, and the MDF board would have 100-150mm frame under it to prevent bowing

it also seems to me that the calculated 150 mikron expansion for 120mm sheet of SS over 10 degrees is quite a ways off, it would much less, like 10x less, but then again, I've welded some SS that moved around a LOT more than simple thermal expansion/contraction would make you think, probably due to internal stresses left by manufacturing process
 
I don't think the moisture effect would follow day/night cycle, its too quick
Joking are ye? Take a sheet of MDF and a window spray bottle of ignorant water and a warm-air gun.

Make it bend to yer will off alternate side wetting and drying in a matter of mere minutes, each go. MDF is "MEDIUM" density compressed fuzz. Wood-fiber analog of animal-wool fibre hard FELT.

It ain't near-as-dammit functional-equivalant of pure Lignin "Benelex":

BENELEX Trademark - Registration Number 0830876 - Serial Number 72237140 :: Justia Trademarks

ARBORON BENELEX BENELUX MELAMINE CLAD THERMOSET LAMINATES PHENOLICS FROM COMPLEX PLASTICS TOLL FREE 1-888-PLASTIK 1-800-363-2870

As to "TEN degrees? Never listened to a metal roof start to have conversational ARGUMENTS with itself as the sun comes up or goes down?

130 F swing - whatever that comes out to in big-fat elementary-school training wheels degrees be - is not uncommon.
 
yes it would be 10x less, because the table is 1200mm. the mdf is epoxy sealed on all sides and the moisture effect would bend it the other way anyway. actually i added some flexibilizer to the epoxy, but the sample i kept was elastic for a couple of weeks but then hardened quite a bit for some reason.

well, obviously this is happening. is anybody able to do the math?

originally i wanted to use a 2mm sheet that could then lay flat by itself. but i found out i could live from that money for a month.

edit: i just saw the thread on circle calculation below. so: chord=1200mm, segment height=20mm, resulting arc length=1200.9mm. just as i thought, it shouldnt be happening, we have 0.1mm at best (i doubt the temp. diff. really is 25°c). where am i wrong?
 
yes it would be 10x less, because the table is 1200mm. the mdf is epoxy sealed on all sides and the moisture effect would bend it the other way anyway. actually i added some flexibilizer to the epoxy, but the sample i kept was elastic for a couple of weeks but then hardened quite a bit for some reason.

well, obviously this is happening. is anybody able to do the math?

originally i wanted to use a 2mm sheet that could then lay flat by itself. but i found out i could live from that money for a month.

I've been holding out until I could make sure I had the SPACE for at least a one-inch top, and re-inforced over a stout frame, even so.

But in ignorant A-36, or "don't care" alloy, and NOT costly stainless. Ordinary heavy steel plate just isn't all that costly. Go fetch, then fab safely and well with no-shake legs is another matter.

2mm? That ain't no "work" bench. Its a kitchen wheelie cart top so as to make clean up of food spills easier.

I've used FIVE mm SS to make heat shields for the oven in Hong Kong or fascia for rack-mount servers. Stiffer and easier to drill and such than the thin sh*t metal.

"live on that money for a month?"

You need to mebbe SMUGGLE in some ignorant steel, 'coz for damned sure you are paying over the mark.. or sleeping rough and living outta trash cans?

Steel is cheap, if I failed to mention that.

It's our scarce TIME as costs the very Earth.

Just try buying some "extra time" off eBay and stashing it in a SwissPost account for a future day's use.

:D
 
that sheet was $150.

Well.. it was probably finished drop-dead gorgeous as if for a high-end kitchen or hospital lab, too, wasn't it?

Lot cheaper in "mill finish". FIRST time I bought in HKG, I went in the door out along "Reclamation Street, Kowloon side, with the Yew Ass of Ay "Bigbox" mentality, took away a "stock sized" sheet, taxi all the way home. Then had to lay it across the bathtub on plywood and cut Stainless with a damned saber saw.

Tedious. Painful!

But I had learnt sumthin' useful off the exercise.

Thereafter? Strolled in with my finished dimensions. Lads sheared it to right tight spec, deburred with a file swipe, laid it onto the freight scale, charged me by the fraction of a metric ton for only what I was taking out the door, no fee for the shear work.

Then also taped the edges, wrapped it in newsprint, added a poly-twine carry handle so I could save taxi fare.

How so? By pretending it was just a repo art print to hang on the wall.

MTR rail, y'see has a policy. They haul PEOPLE, never FREIGHT.

Then again? Hong Kong is still "Chinese" only when they want to be.

"Rules" are for Communists, and they are anything BUT that!

:)
 
thermite, I'm guessing no one is spraying water on that mdf board, air humidity changes would affect such things, significant changes would be noticeable between winter/summer, but not so much from day/night, especially if he says that the mdf is epoxy covered

one effect I forgot to mention is that epoxy shrinks when curing, the effect is significant, I've made quite a few composite sheets (glass fiber, carbon, kevlar) in the past, and if you don't maintain laminate symmetry in the sheet, for instance, you laminate carbon-glass-glass, the sheet will warp, if you laminate carbon-glass-carbon, the sheet will stay flat after curing, so in case of stainless-epoxy-mdf, once it cures, it will already tension the mdf side of the stainless and will want to bow the SS upwards, and the temperature will just be an added help doing that

I made a mistake earlier writing 150 mikron differential expansion for 120mm, should have been 120cm

that being said, to achieve a 20mm rise in the middle, considering the bow is regular, and width to stay the same 1200mm, the length of the bow would need to be aprox 1200.900mm, growth of nearly 900 mikron, which is significantly more than SS would grow alone with a 10 degree rise in temperature, which would be: 1200.200mm, and a LOT more if you count in that the MDF will also grow, then the growth difference would be only 50 mikron of the required 900 to achieve that 20mm bow
 
Thermite is 100% correct. Any wood/steel construction for use outside does not work hard fastened together. The wood, MDF or whatever just moves around with temperature and moisture way too much. You need another plan. Don't fight physics......you will lose.
 
thermite, I'm guessing no one is spraying water on that mdf board, air humidity changes would affect such things, significant changes would be noticeable between winter/summer, but not so much from day/night, especially if he says that the mdf is epoxy covered

one effect I forgot to mention is that epoxy shrinks when curing, the effect is significant, I've made quite a few composite sheets (glass fiber, carbon, kevlar) in the past, and if you don't maintain laminate symmetry in the sheet, for instance, you laminate carbon-glass-glass, the sheet will warp, if you laminate carbon-glass-carbon, the sheet will stay flat after curing, so in case of stainless-epoxy-mdf, once it cures, it will already tension the mdf side of the stainless and will want to bow the SS upwards, and the temperature will just be an added help doing that

I made a mistake earlier writing 150 mikron differential expansion for 120mm, should have been 120cm

that being said, to achieve a 20mm rise in the middle, considering the bow is regular, and width to stay the same 1200mm, the length of the bow would need to be aprox 1200.900mm, growth of nearly 900 mikron, which is significantly more than SS would grow alone with a 10 degree rise in temperature, which would be: 1200.200mm, and a LOT more if you count in that the MDF will also grow, then the growth difference would be only 50 mikron of the required 900 to achieve that 20mm bow

Y'all are so lazy as to write in English but measure otherwise, how about shifting to percentages of length and width?

Can't be bothered to read a post if one has to stop and do funny-number conversions every sentance.

Getting the impression this dog wuddna been FUBAR to begin with y'all had but paid MORE attention to the inherent properties of the materials and LESS attention to numbers rationalizing the f**k-up yah haven't yet done SQRT-FA to correct or do-over, properly, next go!

Be a long while before masturbating numbers after the fact as theory fixes an actual benchtop, physically.

Just the way some shit IS.
 
I'm terribly sorry for the confusion, here are more traditional units and corresponding numbers:
table length - 2 and 31/50 cubits
the total expansion of said SS sheet over 50F would be 8 and a half twips, compared to MDFs only 6 and 3/8 twips
the resulting bow of is just over 1/3 nails high
 
Another factor to consider. Steel top is in sun. MDF is in shade below SS. Epoxy and MDF have low thermal conductivities.

Long winded way of saying. MDF is COOLER than SS. This increases expansion differences.

Now some one who cares can do a calc to see how much cooler the MDF would need to be to lead to the observed bowing. Or another interested type could build and FEA model and run it in ANSYS or even Solidworks.
 
I'm terribly sorry for the confusion, here are more traditional units and corresponding numbers:
table length - 2 and 31/50 cubits
the total expansion of said SS sheet over 50F would be 8 and a half twips, compared to MDFs only 6 and 3/8 twips
the resulting bow of is just over 1/3 nails high

I prefer to work in "beard-seconds" for smaller figures.

On topic, has anyone yet considered possible swelling of the MDF-epoxy layer as a culprit? Maybe the binder chemically reacted with the epoxy and made it swell on that side. Since it couldn't move in shear relative to the bonded stainless it just bowed?

Just to make sure I understand, it is bowed with the high point in the middle, or the low point in the middle?
 
to clarify: the table arches up in the middle, its convex. that means the ss top is expanding more that the mdf. at night its flat, in the sun it bows up. the mdf gets warm, the ss is warmer. and again, the mdf is epoxy sealed.

im convinced the low modulus epoxy cant bend the table. it would have to stretch the ss, right?

my problem actually is that i cant bolt the table down to the frame, my ss sheat would shear off. maybe it will shear off anyway. i have to cover the table when not using it and that has no be done in a way so wind doesnt blow the cover away, so its a pain.

the only solution i see is to make two expansion slits in the ss, but i have to think of the best way to do this.

but i still have a strong desire to understand why its happening. the above calculations apparently are wrong by a factor of 10.

edit: i tried to look up movement of mdf caused by humidity and find the subject pretty complicated:
(PDF) Linear expansion and thickness swell of MDF as a function of panel density and sorption state
 
I misunderstood the direction of bowing in my initial reply, I thought the MDF was the expanding component and so the surface was bowing concave with the SS up. There was also no mention of the entire MDF board being coated [as always, information comes piecemeal].

With what I now understand, what I'd suggest is removing the SS sheet from the current MDF board (if need be by routing off strips at a time to avoid bending the metal), prepare a new MDF board (including epoxy sealing all sides), then at whatever temperature is best bonding the old stainless sheet to the bottom with a more elastic adhesive like a polyurethane.

Then, bond a second SS sheet to the top, at least 1.5mm thick. You'd mentioned cost of the original 0.5mm sheet as high, but I have to believe there are sources in neighboring regions that would have offcuts of a thicker sheet at reasonable prices, even if scratched. A thicker sheet will be much more robust, won't dent as quickly as the thin sheet, and will have margin for refinishing if it gets too marked up. But even the 0.5mm sheet will have the tensile properties to balance the top sheet at whatever temperatures you'll use the table in.

I would want to use 3mm material, but I overdo everything...

Perhaps some of the PM members in Europe could suggest sources for a thicker/cheaper SS sheet that could be shipped to CH?
 








 
Back
Top