The First Time I Saw 795 Fail
I'm the office administrator for a 70-person engineering consulting firm. I manage all our building maintenance and supply ordering—roughly $150,000 annually across 8-10 vendors. When a project manager comes to me with a problem, it's usually urgent, and it's usually something that shouldn't have happened.
So when the head of our lab came to me holding a piece of aluminum with a bead of silicone peeling off like a bad sunburn, I knew this was going to be one of those days. "This Dow Corning 795 cost us $400 in labor and materials," he said. "And it failed."
He was angry. I was skeptical. We'd been using 795 for window glazing and small-scale sealing for years. It's basically the gold standard for structural silicone in certain applications. But here it was, failing. And my first thought—which I'm guessing is the same one you're having if you've ever had this happen—was: Did we get a bad batch?
Bad batch. It's the go-to excuse for anyone who has to explain a product failure. And in 99% of cases, it's wrong. This time was no exception.
The Surface Problem
After five years of managing these relationships and dealing with the fallout of failed projects, I've come to believe that material failure is rarely about the material. It's about the substrate.
Lab guy showed me the failed assembly. The aluminum had been cleaned. I could see the wipe marks—some solvent, probably isopropyl alcohol. The 795 was applied, tooled, and left to cure for 48 hours. Classic procedure. Textbook. And it still failed.
It took me 2 years and about 40 failed repairs to understand that clean isn't the same as prepared. (Honestly, I wish I'd learned that lesson before it cost us money.)
The problem: aluminum anodizing. You can wipe it down with alcohol until your arm gets tired. You can scuff it lightly (which lab guy hadn't done). But the surface has to be chemically compatible with the sealant. Dow Corning 795 is designed for structural glazing—it's meant to bond to glass. It can work on aluminum. But if the aluminum has any residual lubrication from the extrusion process, or if the anodizing layer is too smooth, the adhesion is mechanical at best.
Why This Matters For Your Sealant
If you've ever sealed a window frame, a skylight, or a tank fitting with 795 and had it fail, here's what you need to know: the silicone wasn't the problem. It was the surface prep—or lack thereof.
The 'it never matters' moment: I knew we should be using a primer on aluminum before applying 795. But the manufacturer's spec sheet—which I had read—said that for some substrates, it's not required. So we skipped it. Once. And that's when the sealant failed. $400 worth of rework later, I can tell you: primer matters.
Reference: Dow Corning 795 Silicone Building Sealant technical data sheet (as of January 2025). For non-porous substrates like anodized aluminum, a primer (like Dow Corning 395 or 1200 OS Prime Coat) is recommended to ensure adhesion. Skipping it is a gamble.
The Deeper Issue: Misunderstanding The Product
Here's the deeper issue most people miss: Dow Corning 795 isn't a 'seal-all' miracle goop. It's a two-component silicone designed for specific expansion and contraction characteristics. Its 'low modulus' means it stays flexible over a wide temperature range—great for glass-to-aluminum joints that experience thermal cycling. But 'low modulus' also means it doesn't have the tear strength of a high-performance adhesive like 3145 RTV (which we use for bonding, not sealing). And it's not designed for submerged applications or constant water pressure.
Our lab guy was using it on a fitting that would see intermittent water spray. The silicone was doing its job—it was flexible. But the mechanical bond to the aluminum wasn't strong enough to handle the water pressure. It was a mismatch of material to application.
Take it from someone who's made this mistake: You need to ask 'what is this sealant actually for?' before you pick a tube. A silicone that's great for weatherproofing windows may be a terrible choice for a water manifold.
The Cost Of Getting It Wrong
Let's talk about the price of a bad sealant choice. I don't have hard data on industry-wide failure rates. But based on our experience—and I've seen this happen 4 times in 5 years—the cost multiplies fast.
Direct costs:
- Failed sealant removal: 1-2 hours of labor at $75/hour = $75-150
- Surface re-preparation: another hour, $75
- Re-application with proper primer: $25-50 in materials (including primer)
- Total avoidable cost: $175-275.
Indirect costs:
- Lost time on the project (the engineer had to redo his testing)
- Vendor relationship damage (we nearly switched sealant suppliers)
- Internal reputation damage (lab guy made me look bad to the VP)
I only believed that surface prep was the real villain after ignoring it and paying $400 for a rework. The 'budget' choice—skipping primer—cost us way more than buying a $15 bottle of 1200 OS Prime Coat.
The Real-World And Tank O-Ring Angles
This same thinking applies to other sealing situations. If you're dealing with a tank o-ring seal, the issue isn't 'what rubber is best?' It's 'what is the chemical and thermal exposure?' Nitrile rubber (Buna-N) is great for oil and fuel up to 100°C. But it will swell and fail in ozone or extreme heat. Silicone (VMQ) handles high and low temperatures but has poor tear strength and can swell in oil. The wrong choice here costs you a tank failure and potentially a chemical spill.
Similarly, with nitrile rubber tubing, the question isn't 'is nitrile good?' It's 'what are you pumping through it?' If it's a hydrocarbon-based fluid, Nitrile is a solid choice—it's fuel-resistant. But if you're running hot water or steam, you'd be much better off with EPDM or silicone. The cost of a burst tube at 60 PSI is not just the tube—it's cleanup, downtime, and risk of equipment damage.
PTFE Coating vs Teflon Coating: The Marketing Trap
This last one is a trap I see all the time in our supply orders. 'PTFE coating vs Teflon coating' is a distinction without a difference. Teflon is a brand name owned by Chemours; PTFE (polytetrafluoroethylene) is the chemical name. There is no functional difference in the coating itself. What does matter is:
- The application method: Is it a liquid spray applied and baked, or a powder coat? This affects thickness and uniformity.
- The surface preparation: For PTFE to bond to metal, the metal needs to be roughened—usually by grit blasting. A smooth surface will cause the coating to peel.
- The performance: PTFE coating is non-stick and chemically inert. But it's soft. Hard particles can score or scratch it, exposing the metal underneath.
If you're specifying a 'PTFE coating,' you're basically saying 'I want a non-stick, chemically inert surface.' If you're asking for 'Teflon coating,' you're asking for the same thing, just from a specific brand. The real decision is: do you need non-stick, or do you need corrosion resistance? For the latter, a ceramic or epoxy coating might be a better choice.
The Solution (Short, Because It's Obvious)
Here's the thing: once you understand the root cause, the solution is almost boring.
For sealant failure:
- It's almost always surface prep. Clean, scuff, prime. Don't skip steps.
- Read the spec sheet. Actually read it. The manufacturer tells you the limitations.
For material selection:
- Stop asking 'which is best?' Start asking 'which is best for this application?'
- Temperature, pressure, chemical exposure, UV exposure—these are the variables, not brand name.
For administrative purchasing:
- Create a simple checklist for any sealing or coating job: What is the environment? What is the expected life? What surface prep is required?
- Don't let engineers or lab guys skip the prep step just because they're in a hurry. The money you save on rework is real.
As of January 2025, this is how I handle every sealing request that comes across my desk. It saves us time, money, and a lot of awkward conversations with the VP. (Not that we ever have those anymore.)