Improve icing physics on aircraft

have you tried icing yet?

I just noticed something odd. I had a quick test, then found out that my assistance option for visual only icing had got enabled for some reason. I switch that to ON rather than visual only, flew through icy clouds at -10F, and no ice accumulation. I then dragged the slider all the way across to one, and in the 172 I am gradually seeing it decay back to 0. It looks like the combination of weather, and altitude I was at was not letting ice build. Similarly I noticed that if the temperature is too cold, the plane defrosts itself. I would have thought that whatever ice had accumulated would remain, rather than gradually be removed.

It wasn’t until I climber to about 10,000ft that I started to see ice accumulate naturally, no matter what cloud I was flying through. I also noticed something really odd. If I dragged the ice slider to 1 I saw my airspeed drop a bit. I dragged it to 0, and it dropped again. Back to 1, another drop.

But I am seeing a reduction in lift when adjusting the slider by hand. Turn on the “Sim Forces” debug:


This will turn on visual representation of lift, I believe.

Zero ice

Fully iced

From the external view you can see the plane drop in altitude before the AP recovers. I think the speed drop I was seeing earlier was due to the attitude of the plane changing maybe. When I remove al the ice it noses up sharply, dropping the airspeed.

Other than those blips I see no large changes in airspeed when things settle.

I also looked at drag. I could see that when fully iced, the total weight increases by about 100lbs.



So ice is having an effect on the weight of the plane.


I also found out that when your aircraft is parked at the apron without icing conditions and you move your drone camera to the icing conditions, your aircraft will start to acumulate ice which is also very odd.

Regarding the lift, yes the aircraft are difinitely affected by ice in this sim but the effect is very weak and acumulation of ice is very slow, at least according to me.

For your personally are you satisfied with icing simulation in this sim?

I’m not a pilot, so I can’t answer that with any degree of accuracy or confidence. :slight_smile:

I think there are some aspects of the simulation I don’t fully understand. Like should I not see an increase in drag, or parasitic drag when the airframe is iced over. It does seem to show but as you say the effect is tiny.

Looking at “other_drag”, I see an increase from “0.002” to “0.01”. I have no way of knowing whether that is correct. Similarly I see increases in “cd0”. Similarly small values of reduction in life coefficient.

But no real changes in speed. I would have thought that as the airframe ices over, at the same power level I would need a slightly higher nose attitude to maintain level flight, which would reduce speed, but I don’t think I am seeing that. I can see the plane retrim, but the airspeed recovers.

Exactly, you understand it very well.

I’m a professional pilot and as I wrote earlier, our aircraft in icing conditions can be contaminated by ice within few seconds, when the anti-ice is not turned on. Sometimes even with anti-ice turned on it might not be enough to get rid of ice.

Small unprotected GA aircraft in icing conditions can fall like a rock. In this sim you can fly your Cessna 172 in CB clouds for hours and drink your coffee at the same time.

I think there is no need to provide with more proofs that icing simulation within MSFS is not simulated correctly and does not represent and risk for a virtual pilots at all.


I had a quick poke around in the SDK documentation. There are references to anti-ice dotted about, but only a single icing one.


With this value you can scale up or down the effects of icing on the plane. This will affects the effect of icing on lift and on the weight.

The default value is 1.0 (100% of the effect).

The 172 already has this set to 1…actually scratch that, it doesn’t have it set at all! But its possible that a “default” setting doesn’t actually need to be set at all if you are happy with the default. So the scalar is still set to “1” even if not explicitly listed.

I might have a play later to see if it can be set to 200%, for example.

The accumulation rate was definitely reduced. I’m not sure about the effect though, and whether its impact was reduced. There was a time when your windshield would get iced over in 5-10 seconds.

Just tried this in the steam 172, in the “[FLIGHT_TUNING]” section of “flight_model.cfg”:


I get a huge change in VS when flipping between no ice detected, and fully iced.

Level flight at 3000ft @ 138kts.


Flip to fully iced, and I’m in a 1500fpm decent:


Trim could not adjust fast enough. When I pulled back on the yoke I stalled, and went into a spin.

The amount of ice I can now load up has increased as well, so I’m pretty sure that scalar is doing its job!

No ice:

Fully iced:

Lift has dropped right off, with huge changes in the graphical representation.

No ice:


Fully iced:

I don’t think we can affect the rate of accumulation, but we can certainly mess with it’s effects.


That is very nice to see, that this sim is capable of simulating this! That is exactly What I want to see when flying with unprotected aircraft in severe icing enviroment.

Otherwise there is no point of avoiding icing enviroment when it cant really end your flight. No need to turn anti-ice swtiches ON either.

Thank you for your time and testing which you have done!

However guestion remains, how is that possible, that we dont have it like this in the “normal” version of the sim?

I would suggest to also add some slider or options for realistic, normal and light icing.


I suppose it would be one of those things difficult to simulate accurately, as no one in their right mind would voluntarily fly into those conditions to find out.

All I did was change a 1 to a 3, assuming for a moment that the setting being missing equates to 1. I have no idea whether that is realistic, as the numbers are effectively arbitrary, but it certainly gives you a new found respect for icing conditions.

Without being able to adjust the rate of accumulation, all you can do is make that number larger so that’s it’s effects are felt sooner. But that would lead to a disconnect between the visual icing, and the icing effects. Too high a number, and you might have very little visual icing yet fall out of the sky like a rock.

1 Like

You are absouletly correct, however you can find yourself involuntarily in icing conditions, it happend to many aviators and it will happen to many due to the poor weather briefing, overconfidence and so on.

Simulator should simulate all hazards and danger which can happen to you during flight in my opinion.

I dont like to compare but xplane got this right, they find the way how to simulate it, so you have to think twice before flying into the icing enviroment.


That’s good to know!

I’m going to have a play around with

on the C208.

At least with this knowledge we can all tweak our own aircraft so that icing is actually a hazard. (But we shouldn’t have to, it should be correct ‘out of the box’)

I’m going to tweak the C208 icing scalar so that as a bare minimum it cannot stay airbourne with icing. I’ll do this as shown below:

  • Full power[2397ftlbs torque], no anti-icing, clean configuration, min weight
  • AP on
  • Icing conditions with 1 for the aircraft airframe ice accumulation value in dev console
  • Gradually increase the scalar until stall or unable to maintain altitude

I’m really curious to test both the efficacy of the anti-icing system and the effect lowering flaps has within MSFS.

Is anti-icing just magical (making you immune to icing in MSFS), or is it more realistic?


I wonder whether the rate of ice accumulation is affected by several factors like real life (Meteorological conditions, Angle of Attack, Aircraft configuration etc.) or if it is just flat rate?

Because if it’s flat rate, that’s something else we could do with being improved too.


Agree that the evidence is there that icing needs improvement. Not sure we can call it a regression as it was probably never well modelled in the first place!

Great job @hobanagerik :ok_hand:


I was able to get varying amounts of icing as I played with temperature. Watch the numbers on the slider for icing, vary the temperature, and see how the rate of increase changes. I couldn’t find a way to pin that readout so every time you click away from it to change the weather it closes, but I’m pretty sure the rate change as I made changes.

I may have been doing something wrong, but I couldn’t get icing to start happening until I was about 1500ft up, no matter what temperature, and cloud height I set.

1 Like

This is how xplane 12 is simulating icing.


Yep. Heavy icing not compatible with flight.

That stall horn though! Thought it was a parrot at first.


:joy: you made my day

nah it’s one of those desert frog things. A tiny angry squeaking Frog 🐸 | Super Cute Animals - BBC - YouTube


FYI, I’ve opened a topic during SU12 beta about the regression of the aerodynamic impact. It has become a bug-report topic and now marked as “bug-logged”. Let’s hope they will get this fixed soon enough …
REGRESSION: Icing has no aerodynamic impact

Found out about this excellent thread thanks to @AlpineB4652 :+1:


With the unbearable humid summer hell finally over now I am getting really excited for autumn and the romantic winter time. :smiley:

On some videos I have seem how awesome and super realistic the snow is falling on X-Plane aircraft and accumulating on it´s wings and upper part of the hull.
Important to have this too in FS2020 (or FS2024 please).
Plus I would love to see thick ice accumulation on the slats and the elevators and rudder and other parts of the airplane (not just colouring the slats in white but real half-transparent 3D ice when aircraft anti ice is switched off).

In it´s early days after release FS2020 had absolute awesome icing effects with the airplanes completely freezing over like it was the Asobo lead developer´s Peugeot in an icy January night, and also the airplane windows went completely frozen with the view getting refracted when looking through the frozen window until window heating is switched on.

The icing effects where already truly good looking and it was a joy to unfreeze an airplane with cold and dark cockpit parking in thick snow, have the completely frozen over windows defrosting with the window heating, calling the trucks with the anti ice fluid (this looked really good in GSX)…
Until a very dark day came when literally all winter and icing effects -------> 100% removed from the simulator.

Well I want this day reversed please, and also beautyful snow accumulation and freezing and ice effects really brought back into the sim to shine, and make awesome exciting winter ops possible.


It would be nice to have the icing improved finally. Flight simming is not only about the eye-candy.