[ATM] RE: Silicone Failure

[Mark Holm <mdholm@telerama.com>]

I think if you set down to calculate the stresses, you would find that a 
properly adhered and cured silicone should not fail.  Even though 
silicone is not the strongest stuff around, the stresses are not going 
to turn out that high.  Besides, you have the Steward (U of Arizona) 
mirror lab experience to go on saying that, properly done, a whopping 
big mirror can be supported this way.

The failure in this case is most likely a matter of incorrect material 
choice, or incorrect technique.

Here are some quotes from the D-C product sheet for D-C 832

• 100 percent silicone RTV rubber
• Superior unprimed adhesion to a
wide variety of surfaces, including
glass, metal, masonry, concrete,
most rubber and plastics, including
ABS, PVC and polycarbonate
• Low odor
• Neutral (alcohol) cure
• Will not corrode metal or other
corrosion-sensitive materials
• Use temperatures: -67 to 300°F
(-55 to 149°C); intermittent to
350°F (177°C)
• Forms flexible, long-lasting seal
• Available in off-white, black and
gray

Apply to clean, dry, frost-free surfaces.
Typically, no primer is necessary; however,
for best adhesion to concrete, use a
primer such as Dow Corning ® 1205 Primer.
Dow Corning 832 Multi-Surface Adhesive
Sealant will dry to the touch in 20 minutes.

• Not intended for food-contact
applications
• Requires atmospheric moisture to cure
• Not for use in submerged structural
or adhesive applications
• Not designed for high-movement
joints
• Not for application to oily or dirty
surfaces
• Most paints will not adhere to sealant

The full data sheet can be had at 
http://www.dowcorning.com/applications/product_finder/pf_products.asp?l1=002&prod=M0001461&type=MRKT
click the Product Data Sheet link on the right hand side.  They require 
your email address before letting you see it, but I am not particularly 
worried about Dow Corning spam.  Even if they do send me some email, it 
would be a welcome variation on the v1Agr@ spam.


The following is quoted from  "Dow Corning® brand Primers, Prime Coats 
and Adhesion Promoters"

SURFACE PREPARATION
The active ingredients must thoroughly wet-out and coat the
bonding surfaces. Mild abrasion, solvent cleaning, plasma,
corona discharge and other pre-treatments have been used
to clean and enhance surface reactivity to bonding. In
general, light surface abrasion is recommended whenever
possible, because it promotes good cleaning and increases
the surface area for bonding. Surfaces should be cleaned
and/or degreased with Dow Corning ® brand OS Fluids,
naphtha, mineral spirits, methyl ethyl ketone (MEK) or
other suitable solvents that will remove oils and other contaminants
that may be present. A final surface wipe with acetone
or IPA may also be helpful.

I think the "light surface abrasion" recommendation is very likely a 
good one.  Hit the metal with some 320 SiC paper, then clean it with 
solvent and do the gluing as soon as the solvent dries.  (Use a clean 
grade of solvent, definitely not gasoline, kerosine or turpentine.)

You might look into Dow Corning P5200 primer.  It is formulated in a 
rather exotic solvent, "Dow Corning ® OS-20 is a 1-cSt ozone-safe 
volatile methylsiloxane fluid."  I don't know anything about the 
toxicity aspects of this stuff.  Most of the methyl siloxanes are pretty 
low toxicity, but most of them are obviously a lot higher molecular 
weight that this.  It isn't quite as flammable as many solvents.  It is 
probably expensive.  You wouldn't need much though.

-- 
Mark Holm
mdholm@telerama.com

_______________________________________________
ATM mailing list http://www.atmlist.net/