6

This is a quote listed everywhere in aviation safety publications, but not in a consistent form.

From http://www.aopa.org/-/media/Files/AOPA/Home/Pilot%20Resources/ASI/Safety%20Advisors/sa11.pdf:

Frost, snow, and ice accumulations (on the leading edge or upper surface of the wing) no thicker or rougher than a piece of coarse sandpaper can reduce lift by 30 percent and increase drag up to 40 percent.

http://www.tc.gc.ca/eng/civilaviation/publications/tp10643-chapter1-concept-1119.htm:

Test data indicates that during takeoff, frost, ice or snow formations having a thickness and surface roughness similar to medium or coarse sandpaper, on the leading edge and upper surface of a wing, can reduce wing lift by as much as 30% and increase drag by 40%.

Another publication by Transport Canada removes the "during takeoff" caveat http://www.tc.gc.ca/media/documents/ca-publications/AIM-2013-1_ENG.pdf (p. 423):

Test data indicate that frost, ice or snow formations having a thickness and surface roughness similar to medium or coarse sandpaper, on the leading edge and upper surface of a wing, can reduce wing lift by as much as 30% and increase drag by 40%.

This publication removes the restriction to the leading edge and upper surface of the wing http://www.jumpjet.info/Emergency-Preparedness/Disaster-Mitigation/Climate/Aircraft_Icing.pdf

tests have shown that icing no thicker or rougher than a piece of coarse sandpaper can reduce lift by 30% and increase drag by 40%

Is it true that icing no thicker or rougher than a piece of coarse sandpaper can reduce lift by 30% and increase drag by 40%.

Is it true even when the ice is restricted to the upper surface and leading edge of the wing?

Is it true at times other than takeoff?

  • do a google search for "plane crash icing", there are plenty of examples of planes crashed by ice on the wings – ratchet freak Sep 09 '13 at 18:58
  • 2
    No. I know planes crash from icing. –  Sep 09 '13 at 19:23
  • *Is it true even when the ice is restricted to the upper surface and leading edge of the wing?* - Part of what causes the lift is the shape of the leading edge forcing the airflow in a certian pattern. If that gets disrupted it can reduce the efficency of the wing yes. I am not good enough with the math to prove 30% for 1/16th of an inch though I suspect that is about right. – Chad Sep 09 '13 at 20:22
  • I don't see a notable claim here. Aircraft crashes due to icing are a well-known phenomenon, and I don't see any notable claim that what is written in these papers is untrue. – DJClayworth Sep 09 '13 at 20:23
  • 2
    @DJClayworth The papers themselves are the notable claim, that the effect happens as stated. Is there evidence for or against those claims? – ChrisW Sep 09 '13 at 20:47
  • Is there a notable claim that this doesn't happen? – DJClayworth Sep 09 '13 at 20:50
  • "Mulherin, ND, RB Haehnel, JF Jones (1998) Toward developing a standard shear test fro ice adhesion. Proceedings, 8th International Workshop on Atmospheric Icing Structures, Reykjavik, Iceland, June 8-11, 1998. IWAIS ’98" – DJClayworth Sep 09 '13 at 20:54
  • 1
    Djclayworth we don't require notable claims counter to the notable claim being asked about. –  Sep 09 '13 at 20:58
  • 1
    Please explain to me why you doubt what is written in the publications you reference? – DJClayworth Sep 09 '13 at 21:01
  • 1
    I don't need to. But, primarily because they talk about some tests that have been done establishing the 30 and 40% figures, but don't properly reference the source. –  Sep 09 '13 at 21:07
  • @DJClayworth I'm not sure how your "Toward developing a standard shear test" reference relates to the question: I cannot find a transcript or an abstract for it, but documents which cite it imply that it may be about how to test ice-shedding coatings such as silicone? – ChrisW Sep 09 '13 at 23:11
  • related: a rough surface vs a polished one is crucial for [swing bowling in cricket](http://en.wikipedia.org/wiki/Swing_bowling), – ratchet freak Oct 14 '13 at 12:13

2 Answers2

10

The phenomenon is discussed in this paper which says on page 24,

Wind tunnel and flight test data show that modern high performance airfoils stall at lower AOA when the wing leading edge is contaminated.
Roughness along the wing leading edge equivalent to #40 grit sandpaper can cause the stall AOA to reduce by over 5°, with a corresponding loss of maximum lift capability.

Changing the angle of attack has a large effect on lift. Page 17 of the paper shows that the change in AOA can even become catastrophic.

It's especially true when there's a high AOA already, for example during takeoff.

If it's accurate, Figure 9 of http://allstar.fiu.edu/aero/airflylvl3.htm suggests that a 5° change in AOA corresponds to approximately 30% or 40% lift.

As for drag, http://www.sandford.org/gandercrash/investigations/majority_report/html/_appendixc.shtml explains,

Ice contamination of an aircraft wing also has a significant detrimental effect on the aircraft's total drag, that is, the force which resists the aircraft's forward motion through the air. The total drag has two components, parasite drag and induced drag. Induced drag is that drag which is produced by the generation of lift. Induced drag increases as the angle of attack increases. Therefore, since a contaminated wing must fly at a higher angle of attack at a given airspeed to produce the required lift, the induced drag generated at that airspeed will be higher than the induced drag of an uncontaminated wing. Furthermore, since ice contamination causes the airflow to separate earlier from the upper surface of the wing, it results in a higher induced drag value at any angle of attack. The increase in parasite drag as a result of ice contamination is small in comparison to the increase in induced drag.

This implies that the two figures (life and drag) are related: because it reduces lift by 30%, then the pilot changes attitude to compensate, therefore the drag increases.

Because of the Conclusions at the end of http://allstar.fiu.edu/aero/airflylvl3.htm it doesn't surprise me that the change in drag is proportional to the change in lift; and the second-last sentence quoted above may explain why it's slightly higher (e.g. 40% instead of 30%).

This isn't a good answer IMO, but I hope it's better than none: it explains "why", and shows that the "how much" is at least plausible.

SQB
  • 3,339
  • 2
  • 22
  • 48
ChrisW
  • 26,552
  • 5
  • 108
  • 141
  • Thanks! But it's the 30 and 40% figures that I'm interested in. I don't doubt that icing decreases the critical angle of attack. –  Sep 09 '13 at 21:14
  • @Sancho The graph on page 25 may be relevant, but I can't make it out well enough to understand it. Ditto for subsequent graphs: the vertical axes aren't properly (quantifiably) labelled. Sorry about that. Maybe it's merely an order-of-magnitude approximation, which depends on aircraft type, AOA, etc.: for example it can apparently lead to stalling which I guess means large (larger than 30%) loss of lift. As for the source of the data: it's a Bombardier paper talking about a specific one of their own planes, so it's presumably from their own testing. – ChrisW Sep 09 '13 at 21:35
  • 2
    @Sancho If it's accurate, Figure 9 of http://www.allstar.fiu.edu/aero/airflylvl3.htm suggests that a 5° change in AOA corresponds to approximately 30% or 40% lift. – ChrisW Sep 09 '13 at 21:44
  • 1
    There is some stuff in here that doesn't make sense. (Not unusual in aviation reporting.) It does make sense that roughness can reduce the stall angle of attack, like from 17 to 12 degrees. It does not follow that *at a particular AOA below stall* there is less lift. If the plane is flying straight and level at an AOA of 10 degrees, and encounters icing, lift = weight. The ice does not reduce the lift. Rather it 1) reduces the AOA safety margin and 2) increases the weight (given sufficient accumulation). (And the drag, as you said.) – Mike Dunlavey Sep 11 '13 at 21:15
  • ... cont'd. When planes get iced up in flight, they have to fly faster to carry the extra weight without increasing AOA too close to stall. With the extra drag, this takes extra power, so all the margins are reduced. Planes have crashed on takeoff because the stall AOA is reduced, they are heavy with fuel, and they cannot reach high enough airspeed in the available runway. They may be able to get off the ground, in ground effect, but they cannot rise. – Mike Dunlavey Sep 11 '13 at 21:31
  • @MikeDunlavey I assume it's /not/ the weight of the ice: that's negligible compared to the weight of the plane. Pages 12 through 16 of http://www.sae.org/events/icing/presentations/2007s1tanner.pdf imply that the root cause is turbulence of the airflow over the wing. – ChrisW Sep 11 '13 at 21:33
  • @ChrisW: Right, and slide 25 shows that the roughness does not decrease lift at a given AOA. Rather, it decreases the maximum AOA. – Mike Dunlavey Sep 11 '13 at 21:41
  • 1
    @MikeDunlavey Thanks for your help. On reviewing that graph, I agree with you that that's what that graph is saying. So you're saying that the Gander report appendix is misleading re. the reason for increased drag ... and therefore my theorizing after that quote is wrong too. – ChrisW Sep 11 '13 at 22:31
  • @ChrisW: No problem. If you haven't taken flight lessons, I would encourage it. It's not too expensive, great fun, and interesting, and you'll come to understand this stuff implicitly. – Mike Dunlavey Sep 11 '13 at 23:31
  • @MikeDunlavey From http://www.tc.faa.gov/its/worldpac/techrpt/ar03-65.pdf, ice does reduce the lift coefficient at all angles of attack. It is reduced most drastically at high angles of attack, near the critical angle of attack, but it does get reduced throughout the entire range. –  Sep 12 '13 at 04:05
  • @Sancho: That doc is about accretion ice (which alters airfoil shape), not frost. Figures 38 and 58+ of that pdf show how the ice accumulation mainly reduces the usable angle of attack, above which, as you say, the lift coefficient does suffer. – Mike Dunlavey Sep 13 '13 at 21:38
  • 1
    @MikeDunlavey It *mainly* reduces the usable angle of attack, but it reduces the coefficient of lift at all angles. Even slide 25 of sae.org/events/icing/presentations/2007s1tanner.pdf shows a small decrease in lift coefficient at all angles of attack. It's most noticeable at high angles, though. –  Sep 13 '13 at 21:41
4

This paper (A Silicone-Based Ice-Phobic Coating for Aircraft) is another source of the quote:

In fact, wind tunnel tests have shown that very thin ice sheets can reduce lift by as much as 30% and drag by 40%

It gives a reference:

Mulherin, ND, RB Haehnel, JF Jones (1998) Toward developing a standard shear test fro ice adhesion. Proceedings, 8th International Workshop on Atmospheric Icing Structures, Reykjavik, Iceland, June 8-11, 1998. IWAIS ’98

1998 was too long ago for the papers to be online.

DJClayworth
  • 57,419
  • 26
  • 209
  • 195
  • 1
    It should say "reduce *maximum* lift", which is the result of reducing the maximum AOA. In normal flight, the lift is only a fraction of the maximum lift. – Mike Dunlavey Sep 11 '13 at 21:52
  • @MikeDunlavey What it *should* say depends on the data from the tests it references. –  Sep 13 '13 at 21:41
  • @MikeDunlavey Sancho is right. "Reduce lift by 30%" is perfectly valid if tests were carried out at constant AOA and airspeed. Obviously this would have an effect on normal flight, where a reduction in lift would be countered by an increase in AOA, which would also result in an increase in drag (in addition to the increase in drag from the ice) and...well we all know the end result. – DJClayworth Sep 13 '13 at 23:48
  • 2
    @DJClayworth: Not to split hairs, but that quote is in the introduction to the paper and, loosely speaking, that 30% is correct in the sense of *maximum* or *available* lift. What's more, if you look up the reference as [_here_](http://www.scientiasc.com.br/img/arquivo/Arquivo-39.pdf) and [_here_](http://www.silicone-polymers.co.uk/pdfMaster/Silicone%20Ice%20Phobic%20Coatings.pdf) maybe I'm blind but I saw no such claim. I think slide 25 of ChrisW's [_sae.org reference_](http://www.sae.org/events/icing/presentations/2007s1tanner.pdf) was squarely addressing the question. – Mike Dunlavey Sep 14 '13 at 01:12
  • 1) Yes, it's in the introduction 2) I've already said that a reduction in lift is what they claim ,and I see no reason to doubt it 3) I have no idea what the two links you quote are for, except that they both reference the original paper too. The fact that they don't make the same claim may simply mean they are talking about a different aspect of the paper 4) The sae paper is obviously in the same area, but since it doesn't make the same claim why is it relevant? – DJClayworth Sep 14 '13 at 02:44
  • `1998 was too long ago for the papers to be online.` I have issues with this statement, as I regularly find papers that pre-date the '80s. The problem in this case is that the article was a conference paper, rather than a journal paper, and the former tend to be less available online. – Federico Apr 25 '16 at 12:53