Menu Sign In Contact FAQ
Welcome to our forums

TKS ice protection for a TB20

It turns out the installation manual calls up for some PTFE adhesive tape to be taped onto the wing where that particular TKS panel goes, plus nutplates to be installed from the inside (which is an absolute bastard of a job), so that panel is not all that well attached with the 3204 sealant, and can be removed for the stall warner service if necessary.

This might be relevant but is not feasible to do legally on any meaningful timescale.

Shoreham EGKA, United Kingdom

About a week to go. Panels provisionally in place, awaiting the massive gluing session. I have not been able to get the Chemseal 3201 sealant at all, needed for the TKS panel edge fillets. The STC doesn’t call it up but it does a better job.

Shoreham EGKA, United Kingdom

All TKS panels installed. It was a 10hr marathon job. It went well but only because I spent several days shaping the panel edges; half that time was overcoming really unnecessary and irritating manufacturing defects on the back of some of the panels. In the foreground is a wooden fixture used to shape the large wing panels.

We used up all 15 sempacks of the 3204 sealant. The panel over the stall warner is not fully fixed yet because we need to do something there and in any case it isn’t fixed with sealant as such; it is held by 8 screws and is sealed only with a fillet around the edge. The fillet will be applied to all panels late next week when the main job sealant has cured (at the low temps we have, despite heating in the hangar 24/7, this will take about a week).

Due to some reports of a speed loss up to 7kt (outrageous IMHO) great care was taken to get a really good fit against the wing, so the fillet needs to be only the thickness of the material at the edge of the panel – about 1mm.

On the lower cowling you can see the circular window for the ice lamp. Per STC, this was supposed to be just a huge 52mm hole. I put some toughened glass in there (£5 on Ebay) and it looks much better.

The reason some panels are not taped is because they are riveted. Obviously there are no rivets in the fuel tank areas

I have the old prop TKS parts for sale – the tank+pump, its mounting bracket, and the cockpit control unit. I have just heard from Socata that the first item is €6000+VAT. I reckon, looking at other similar items, the control unit is €1000-2000.. I reckon I would sell the lot for €2000. The buyer would then need to source the propeller metalwork.

Shoreham EGKA, United Kingdom

Some random pics from the installation.

Shaping the detail of the panel edges to fit the wing surface exactly

Stall warner cut-out

The elevator panels are curiously not mounted symmetrically even though the elevator aerofoil is symmetrical – this was checked extremely carefully and seems to be really correct although nobody has any idea why

The elevator panels are held by loads of rivets and don’t really need much sealant except for vibration damping

The nylon tubing and the fittings… those olives are £12 each!

The panels, with the areas where the tiny holes are having been masked for some protection

The pneumatic sealant gun, bought on Ebay for £100; new price about £500

The sealant detail – this panel end is not riveted so only the glue is holding it onto the aircraft

One of the many “challenges” in making the panels fit – no comment! (This is some part of the overpressure venting system)

Shoreham EGKA, United Kingdom

A further update – a quick hack of a w&b schedule. This is with full fluid tank, 20kg of junk on the back seat and 20kg of junk in the boot. 2 not too large people in the front and full fuel

Right on MTOW but, as before with the TB20, virtually impossible to load it fore or aft or the envelope. This is about the only way to achieve that (a UK Size 8 lady pilot, RHS empty, two 80kg people in the back seat, and full TKS fluid tank

Interesting they said the TB21 with full TKS was just a 2-seater but actually with the extra 50kg for the turbo it would be a 1-seater!

Shoreham EGKA, United Kingdom

This is how closely the TKS panel edges need to be formed – they have to just touch the skin

The thickness of the panel edge is about 1mm.

Then you do the fillet with the sealant

and the thickness of the fillet goes from the 1mm to around zero, over its ~7mm width.

We used sellotape for the masking because it is really thin so you get no overhang.

The sealant used for the fillet is Chemseal 3204 BLACK B2. I never got the 3201 stuff. CAV report that the 3204 Black is what they use. The STC itself was written about 30 years ago and a lot of it is, ahem…

It will be interesting how much speed loss there is.

Shoreham EGKA, United Kingdom

Wings are finally finished with the fillet

It is beginning to look like a proper plane

Doing the underside fillet is a difficult job because one is bent over and it is hard to get the fillet at an even width.

I am paying the hangar owner £1/hr for the extra 16hrs/day of running the heating, but still the “2hr” PRC sealant takes a week or more to go off because the heater barely manages to do much. This is really a summer job… We left the panels taped on for 2 weeks before removing the tape. This is one of the reasons why the job has over-run by a few weeks.

We had some fun with the concurrent Annual. The landing gear went up and down a few times and then would not come back up. 2 hours later I found the culprit: a Socata factory bodge, where some wires went into a squat switch connector without any strain relief, and predictably broke, inside the insulation. Good job this happened in the hangar and not on a flight; you lose a lot of speed flying with the hear down. On the other squat switch there is some strain relief, in the form of a couple of turns of insulating tape… those connectors are really unsuitable for the job and proper connectors is one of the things you get on a multi million $ plane. There are three switches any of which stop the gear pump: the ~1500psi pressure switch, left squat switch, right squat switch. I carry a 24V soldering iron for this sort of thing

Shoreham EGKA, United Kingdom

The cure time for the sealant seems to be dependent on both temperature and moisture.
By only raising the ambient temperature will not increase the amount of moisture in the air.


Indeed, but I think that for a 2-pack material the moisture is relevant only when it is in a thin layer. In bulk, it can’t get in. So I think that this is mainly relevant to the “pot life” i.e. how long you have to do the job.

77F is +25C is which way warmer than any heated hangar will get to in the winter.

50% RH is also very low for the UK where 99% is more normal 50% is what you get in an air conditioned room.

I reckon this job could be done in 4-6 man-weeks, once you know how and have all the parts in and have inspected them and sorted out anything incorrectly shipped, damaged/secondhand or with manufacturing defects.

Shoreham EGKA, United Kingdom

Shoreham EGKA, United Kingdom
Sign in to add your message

Back to Top