Icing

[ATP_Al]

Can anyone answer the following questions please:

How much more ice would you expect to form on the tailplane than the wing, and why?

I remember reading somewhere that as a rule of thumb, the ice on the tail will be four times thicker than that on the wing. Is that true?

Thanks!

Al

[Old Smokey]

General theory indicates that the thinner is the aerofoil, the more likely it is to pick up ice. As the tailplane is definately a thinner aerofoil than the wing, it should, in accordance with the theory, pick up a larger load of ice. This was certainly so on the Jetstream 31, which was a delightful aircraft until it picked up ice, whereupon it bacame a total bitch. The Propellers, being the thinnest aerofoils of all, are the most ice-prone part of the aircraft.

Regards,

Old Smokey

[safetypee]

ATP_Al it is not the thickness of ice that is the problem, but the shape, location, and type; see Understanding the Stall-recovery Procedure for Turboprop Airplanes in Icing Conditions (http://www.flightsafety.org/pubs/fsd_2005.html). See April 2005, 1.8 mb download, requires Acrobat ver6 +

[lomapaseo]

I would love to know why the thickness of the airfoil has anything to do with ice forming.:confused:

[CaptainSandL]

Loma,

So would I, it goes against logic. I would expect a larger frontal area to catch more ice - and faster, but apparently the opposite is true.

I remember seeing a photograph of a camp either at the N pole or up a mountain somewhere. The guy ropes on the flagpole had inches of ice on them yet the flagpole which was much thicker had none!

Bringing it back to aircraft, the early 737-3/400's had a sharp pointed spinner on the engines. It was soon discovered that these picked up a lot of ice which shed into the engine so they made the spinner round instead.

S & L

[flybubba]

I believe that the icing accumulation on surfaces with smaller dimensions has to do with the airflow characteristics of water droplets. The water droplets are unable to follow an abrupt change in direction due to inertia. It is the same reason bugs smash on your windscreen. In addition the PT-6 inertial ice vanes work the same way.

Regards

[Old Smokey]

I agree with CaptainSandL that it does indeed go against logic, something that I was uneasy about when typing my original response, but observation of thinner tailplanes and propellers at least puts some credit to the conventional wisdom. I think that flybubba might have hit the nail on the head when he said smaller. Now, smaller, not thinner does make sense.

Regards,

Old Smokey

[safetypee]

I quote from BAE SYSTEMS pilots guide ‘Think Ice’

Accretion Efficiency
In general, ice adheres to all forward facing surfaces of the airframe, the accretion rate or catch efficiency being primarily dependent on location and geometry. A relatively large radius aerofoil at moderate or slow airspeeds creates a large pressure wave ahead of the leading edge which forces the air around it, carrying most of the moisture with it. Only droplets sufficiently heavy to overcome this flow will impact the leading edge. Thus, a large chord aerofoil with a blunt leading edge has low ice accretion efficiency.
Conversely, a narrow radius leading edge generates a smaller pressure wave, and so collection efficiency is greater.
In general, the tailplane has a sharper leading edge section and shorter chord than the wing and consequently can accrete ice before it is visible on the wing, and at a greater rate.
The greater mass of rain and drizzle droplets allows them to penetrate the pressure wave, thus giving much higher accretion efficiencies.

[lomapaseo]

Well safetypee

In order to accept your explanation I would have to assume that the bow/leading edge pressure wave is significantly different in pressure (not just size) between thick and thin leading surfaces.

Also under that circumstance the nose radom ice acretion should be much lower than the tailplane, which does seem to go against icing test flight experience, so while agreeing with the effects of droplets inertias vs airflow divergence I don't accept the generalizations of thick vs thin leading edges.

[safetypee]

lomapaseo, not sure that I can elaborate further. My post was a quote not my statement, but I totally agree with it.

From NASA pilots guide to in-flight icing (http://aircrafticing.grc.nasa.gov/courses.html)
“Wings that are thin or have sharp leading edges are more efficient ice collectors. For the reason smaller, thin aerofoils may accrete more ice faster than large thick airfoils. A large transport aircraft will accrete proportionally less ice than a smaller aircraft traversing the same environment”.

The issue “the nose radome ice accretion should be much lower than the tailplane, which does seem to go against icing test flight experience “, is probably one of proportion (as in NASA’s text ‘proportionally’) as the ice accretion depends on the ‘finess ratio’. Therefore you have to compare thick fuselages with ‘thinner’ wings and ‘thinner’ tailplanes, all with different finess ratios; I suspect that in reallity a nose radome to fuselage length has a low finess ‘thin’ ratio in comparison to a wing. Also, I suspect that there are other modifying effects i.e sweepback; the ‘thin’ wing on the Stealth Fighter apparently does not collect ice.

... and now you can tell us about engine spinners, fan baldes etc :-)

[bobrun]

Correct me if I'm wrong, but most current jet transport aircraft do not have tailplane deice....
My question is: if the tailplane is an efficient ice collector, why is the jet transport tailplane not equipped with deice protection?

[411A]

\\Correct me if I'm wrong, but most current jet transport aircraft do not have tailplane deice....
My question is: if the tailplane is an efficient ice collector, why is the jet transport tailplane not equipped with deice protection?\\

Indeed true for older types as well, bobrun, the Lockheed TriStar is but one example.

The answer is...
Inflight icing tests carried out during certification clearly demonstrated that tailplane anti-icing was not required.

[ehwatezedoing]

Inflight icing tests carried out during certification clearly demonstrated that tailplane anti-icing was not required.
So why "exactly" it wasn't required ?
Is it only because the transition done in icing zone by jet transport airplane is rapid enought, its avoid dangerous ice accumulation on their tail !?

[Mad (Flt) Scientist]

Simply, you do tests with artiical ice shapes installed everywhere you don't have deicing - including the tail. If there is no degradation in characteristics which would make you uncertifiable, there's no need to try to deice those places. While the plane may not fly as nicely, as long as it meets the minimum certification standards, it passes.

[411A]

...or, in the case of the TriStar, flying behind the USAF 'flying ice machine' (KC135, as I recall), no appreciable icing accumulation on the tail surfaces.

The proof is in the puddin'.:}

[bookworm]

In order to accept your explanation I would have to assume that the bow/leading edge pressure wave is significantly different in pressure (not just size) between thick and thin leading surfaces.

I don't think so. Imagine two airstreams, one with a large entrained water droplet size, the other with a small water droplet size, flowing towards the same aerofoil. Some of the water droplets entrained in the airflow will deflect around the surface, others will impact it. I think it's intuitive that a greater proportion of the larger droplets will hit the aerofoil than the smaller droplets, because of the greater momentum of the droplets.

Since there are only two dimensions of interest, the radius of the droplet and the radius of curvature of the leading edge of the aerofoil, that also provides the scaling rule: the higher the ratio of droplet-radius/aerofoil-radius, the more rapidly a given shape of ice will accumulate.

Thus if you keep the droplet radius the same and decrease the aerofoil-radius, again a given shape will accumulate faster. Of course there's actually less mass of ice, because it takes less ice to make the same shape on the smaller aerofoil.

[lomapaseo]

safetypee

good, we got rid of that unworkable blunt vs thin leading edge explanation and now we have got down to fineness ratio :)

411A

I never heard of the Tristar and a KC135 Ice machine, but I am familiar with the DC10 and its localized boom to form ice in a single engine inlet, which of course never gets to the tailplane.

I would have some serious doubts about trying to simulate icing on an airframe in anything other than a large cloud.

Mad Scientist I don't see how assumed ice shapes are acceptable means of showing compliance. I was aware that the presumption was continuous ice acretion throughout the whole flight and that the demonstration had to show that any acretion up to a natural shed point had to be shown as not detrimental.

I don't see how one can assume the actual shed vs acretion characteristics without an in-flight test.

Do you have a particular aircraft certification program for reference?

[Mad (Flt) Scientist]

Well, obviously you have to do system demonstrations and/or validation of predictions, but the use of CFD-like predictive methodology to conservatively predict ice shapes for the design point conditions is quite common. I believe that all Bombardier's products have used some combination of natural icing testing (priomarily to demonstrate functioning of the AI systems) and artificial ice shape testing to demonstrate handling compliance.

The codes used to generate the artificial shapes all have to be approved by the relevant airworthiness authorities - there are various codes used by differnt companies and in different countries.

The problem with trying to use natural icing to demonstrate handling compliance is that you can almost never get the right flight conditions - you'll either get favourably good conditions, and may 'pass' a test that you should not, or you'll get conditions beyond the accepted worst-case and fail a test that you should pass.

It's also not (just) accretion up to a shed point. There's a time limit for the exposure based on the severity of the conditions and what you're testing ('failure' cases, or unprotected shapes, or 'turn on' shapes) and there's also a size limitation which is based upon 'accepted' shedding geometry (but which ends up being relatively penalising in practice)

[lomapaseo]

Mad Scientist


Agreed