A_A you can’t always leap to the conclusion that you see no improvement, that it is not a carb heat issue.
If it is severe carb icing it may take a while to clear or it may not clear at all with carb heat.
One can often see a similar effect outside of aviation in mid and southern Europe in an around June. It’s a month where storms carrying large hailstorms occur.
A few years ago, unusually for this area we had one putting dents in cars. At the bottom of the down pipe from our roof gutters a pile of hailstones had congelled.
Despite temperatures reaching the high 20’°s C it took a week for these congeled hailstones to melt and for the downpipe to clear. Not so different from severe icing v carb heat. Admittedly both are abnormal circumstances.
gallois wrote:
If it is severe carb icing it may take a while to clear or it may not clear at all with carb heat.
It would be a spectactularly bad installation if that were the case. Carb heat works by heating up the intake air to a temperature where it remains well above zero. This prevents new ice accretion completely regardless of how humid the air is. And the humidity of the air can never exceed 100%, so designing a system that can cope with all possible carb icing conditions is super easy.
This is fundamentally different from airframe icing which actually STARTS when you have water droplets in the air, and aitframe icing can be several orders of magnitude higher than carb icing.
Why would it be a bad installation ? Venting hot air on solid ice (once it’s accumulated) does not make it melt immediately. And in the worst case (this would be serious oversight by the pilot, but still possible), the carb is fully blocked with ice, so carb heat is basically useless since the hot air can’t get there to melt the ice (also by then the engine has stopped obviously). This is a real deal and can knock your engine out until you descend to higher temps and wait for ice to melt (hopefully before you’re on the ground). The “carb heat” is a poor name btw because there’s generally no heating element around the carb, it’s just like alternate air (using air heated up by the exhaust pipe). I suspect it’s called this way because it’s a counter to carb icing.
The only reliable check that there is no ice left is to get precise MP (or RPM if on the ground), apply carb heat for a few seconds, note MP reduction, remove carb heat, and check the MP is back to the initial value (and not more). As long as you have more MP at the end than at the beginning, you’re not sure to have fully removed the ice.
I posted this video almost 10 years ago, showing the mechanism of ice formation in a carburetor. Note that the ice forms where the fuel is vaporizing, shutting off fuel flow but not throttling the engine. Manifold pressure is not affected by this mechanism. Manifold pressure is affected if the engine starts to quit due to lack of fuel – it then rises due to the reduction of RPM with unchanged throttle position.
There is an extensive discussion at the link and as always people draw their own conclusions. My conclusion is that ice forms first where the fuel vaporizes. The carb’s venturi effect cools the inlet air flow globally, and the fuel evaporation causes an additional local cooling effect that shuts off the fuel jet that is supplying fuel only as a result of a very small pressure differential. When you turn on carb heat the effect is immediate because the amount of ice that needs to be melted to open up the fuel jet is not very much – it does not need to be a big blockage of the inlet tract to almost stop the engine.
In the time since that 2014 discussion I have suffered carb ice once, taxiing on the ground, so this is not something I need to revisit often. 
@Silvaire don’t you live in quite a dry climate?
I live in an area much of which was reclaimed from the sea by Dutch engineers back in the16th century. Known as the Venise Vert there are said to be some 75,000 km of canals criss crossing the area. So you can imagine this area is quite humid.🙂 and as such carb icing is something we are all very aware of here.
On the Rotax 912 we have at the club the carb is always in a flow of heat because of its positioning. The UK LAA still made it necessary for Rotax engines to also have carb heat added in order to gain a permit.
In the French PPL material, carb icing is instead depicted as throttle icing (it happens especially when the throttle is low which creates an additional venturi tube after the carb). When using carb heat the engine can run rough for a few seconds (and it feels like successive blocks of ice are being swallowed by the engine). I suspect in more extreme cases it could be longer, and AFAIR you can lock the entire pipe with ice (at which point you have to wait until it melts by itself). Although this (fortunately) never happened to me, I sometimes need to apply carb heat several times in order to fully get rid of the ice (with a 912).
That’s interesting. Is that a 912 in a certified aircraft? On the 912 in the Super Guépard there isn’t any carb or throttle (obviously a term change since I did my TT, now known as PPL) heat mechanism and so far no icing problems.
Certified (DA20). The training insists substantially on it (the BEA diagram in OP is also in the official PPL theory handbook and there are questions about it in the exam).
I’ve been trained (with actual anecdotes I believe, but it’s already been more than a year) to be aware of icing before completely freezing the engine and having to glide to higher temps (hopefully) and wait for ice to melt by itself (hopefully before being on the ground). Whether ice forms on the fuel nozzle itself or in the throttle, I’m not sure.
This image corresponds to my training, and is the first picture (from wikipedia) that pops when googling it. It shows the ice does not form on the carburetor itself but rather the tube after the fuel vaporizer and on the throttle valve itself (because of the high venturi effect).
@maxbc if I can find an old TT/PPL “manuel de vol” I’ll see what it says and post it if I can. I seem to remember it being much like the picture you posted.I don’t think 4 stroke Rotaxes were around back then. Let alone in certified machines.🙂
The alternatives to Lycoming or Continental IIRC, and mostly for the likes of homebuilt, were, Limbach/VW and JPX and I really can’t remember what the carb/throttle heat situation is.
@Silvaire don’t you live in quite a dry climate?
I am based at an airport in a dry climate, but the conditions can vary between airports I use in the local area, and with time of day. As I write this in the morning, ATIS reports a 1 C spread between temp and dew point at my base. Regardless, I don’t use carb heat on most approaches with my Lycoming. The AFM specifies use “as required” and as with most Lycoming installations I’ve found that carb ice is possible, but not a common occurrence in my use. In contrast the small Continental I flew previously was reputed to be an ice maker and I used carb heat profusely.
