Carb Ice Questions

[Onex107]

When we were making needles we used 1/8 dia. drill rod. We found the needles supplied to vary in diameter. We measured a 2 1/2 and calculated the angle of the taper. When the 2 1/2 was adjusted for WOT we marked the point at the orifice and measured the thickness. Then we adjusted the needle for good idle and marked that point. The object was to get the two "sweet spots" within the slide opening distance. So we increased the angle till that happened. We had a setup on an angle plate on a tool grinder that made it easy to change the angel in minutes. The final taper is very close to the #3 needle supplied. Number #7 was successful.

[berger]

Hi OneX107,

Thank your for your explanation about your systematic way to get the correct needle profile. Great way to go.
On my side, while tuning my carb, back in 2009, I noticed that the different slopes provided by Sonex/AeroVee were not consistent with their numbers, comparisons made. They probably improved their quality control since the AeroInjector tuning subject is no longer a hit on the forums ;-)

As mentioned earlier, I installed a provisional thermal insulation around the gascolator, since the burping syndrome is a well-know issue mentioned by several others Sonex owners, even though I never did on my engine, at least when the --too tight-- engine ran very very hot, during the first 124 hours. After this, yesterday, I went flying within typical air conditions where my engine did show the "unhappiness". Hazy, milky, even. Very high humidity, around 10-12 °C (50-54°F). No "unhappiness" showed up, this time.
It is too early to be victorious since the syndrome is/was typically not appearing on every flight, though.

However, I played a little with the PWR settings and mixture tuning, since the atmospheric conditions were pretty "bad" as far as "unhappiness" is concerned.

I noted some interesting thing: Engine nicely tuned at 2750 rpm, I reduced to 2500. To my surprise, I had to enrich the mixture quite a bit. So I did, to have EGT within Jabiru limits and than, after perhaps one minute of flight within this configuration, I pulled the throttle to idle for a few seconds, without changing anything else. After that, I pushed some power back in to 2500 rpm. Guess what... I had to lean the mixture back. I did this only once, I have to renew this experiment on other flights, but to me, the most logical explanation is that there was some ice on the needle which reduced the fuel flow while going from 2750 to 2500, requiring richening the mixture. When I pulled back to idle, the ice of the needle got scraped off the needle and when I applied 2500 rpm back, the mixture was then too rich because the ice was no longer there.
Once again, these are just assumptions based on the odd observations I've done.

Stay tuned...

Sosthène

[Onex107]

Good work, good observations. Keep it up. I suggest you print a copy of the carb ice graph that is available on the net. Then you can plot the weather conditions of the problem flights. According to that graph, carb ice is possible in almost all combinations of temp and humidity. Just more likely to occur in some more than others. At the time you suspected to shed ice from the needle, did the engine run rough? Even a few drops of water should have effected it. Another thought, my friend Bill Larson is installing a new 3300 in his Sonex and I see the oil cooler and the ribbed oil pan require an air vent for the bottom of the engine. If this is your situation you may not be getting air to the carb as warm as I am in the Onex with the oil cooler on top. An increase in temp without a change in humidity moves the ice point to a less likely location on the graph. I tend to over think these things.

[berger]

Thank your for the idea of plotting the problem flights in the icing probability chart. Might be interesting.

On the latest flight described above, there was no roughness or "unhappiness" at all.
My Waiex has a typical engine installation, as per plans, except for the customized AeroVee air filter. So indeed, there is an air cooling channel underneath the engine.

Since the needle holder has quite some slack at the ball joint connection, and since I noticed that the orifice was laterally worn by the needle wobbling in there, I've recently put some Loctite 111 silicon grease in the ball joint to lessen the wobbling. My thoughts are that this might have introduced some lag between the mixture settings depending on the direction I move the throttle --close vs open.
In the other hand, the grease might well also act like a heat transfer compound, keeping the needle from freezing. Contradictory... Again.

Cheerio,

[flyingbear]

I have had carb ice on a number of engines through the years and I have also had INTAKE ice on a VW powered aircraft. Intake ice, for me, occurred very slowly and dissipated slowly too when carb heat was added- no roughness. I have found that regular carb ice causes roughness when heat is applied.
I am getting ready to fly my ONEX with an Aerovee /Aerocarb and plan to install a cool air system with a carb heat box. I agree with Ellison who states that since guillotine type carbs rarely get carb ice that means that you will rarely die.
I have gotten carb ice mostly around 70--80 degrees OAT but have also gotten it a 32 degrees OAT with regular float bowl carbs.
Dr. Glen Bradley
Thatcher Aircraft

[Wood]

Thanks for this post very good info and comments.
Although carb ice can form anytime of the year this is the prime time for carb icing here. Haven’t been flying piston aircraft for a very long time until this summer. And I had to get back to old habits on this. Sure enough on a flight yesterday I’m glad I did. I always got in the habit of when ever the temp and dew point is within 5-6 degrees of each other to check carb heat every 15-20 minutes or so depending on the temp and dew point spread. Yesterdays flight was no exception and sure enough with temp and dew point close to each other and the RH at 60% as soon as I pulled the carb heat on the RPM drop 300-400 RPM. Defiantly an indication of ice build up. Engine RPM started coming back up to normal after about 10-20 seconds. After that I reduced the time intervals from 15-20 minutes to 10 minute checks.

Stay safe and have fun

Mark