Winter operations / lowest temperature for starting / preheating methods (merged)

I suppose starting a 912, Thielert, or Austro engine by -25 does no more harm than when starting your Mercedes class A or B engine in the same environment. Plenty of those North of 60° North.

Good point. I wouldn’t give a second thought to starting my car engine in cold temperatures. Why is it so different starting an aircraft engine? The air cooled vs water cooled can’t be relevant. So is it just the oil?

One issue is that not all operating clearances in an engine scale with the size of the engine, but thermal contraction does scale directly with the size of the engine. So a physically larger engine with larger cylinder (aluminum inside steel) and larger crankshaft (steel inside aluminum) would see a greater change in cylinder or bearing clearance at a given low temperature.

I suppose starting a 912, Thielert, or Austro engine by -25 does no more harm than when starting your Mercedes class A or B engine in the same environment.

If that were true beyond doubt, I’d be quite relieved. But I understand aviation engines – and the Rotax in especial – are made to a set of compromises quite different from those applied to car engines. Tolerances must be tighter, margins on operating temperatures also; if only because the cooling systems will be less overdimensioned. Ok, that applies to the upper end of the temperature scale and we are now discussing the other side, still I feel there must be less room for variance.

Then again, perhaps the mere existence of this thread has made me over anxious? Actually, it makes me thinking twice about precautions before starting my van’s 3,0 Iveco in -10C that we may soon expect here.

Same goes with bearing clearances, the bigger the bearing, the more clearance. The bigger the cylinder bore, the smaller the alu piston is when cold, more so than the iron cylinder. As I say, a 60mm main bearing will have only 0,03mm less clearance in an alu case at -20 degrees C, 40 degrees lower than +20 deg. C. The mains would be the last place I´d worry about in winter. Vic

the bigger the bearing, the more clearance

eggssactly. The Rotax is amazingly small with its 1211 cc, the bearings must be rather tight.

A couple more points:

The cylinders on modern car engines are typically made from an alloy that matches the thermal contraction of the pistons. A while ago there was a thread about aluminum cylinders being sold for Lycoming engines, and this was one reason I found it interesting. Less variation in cylinder clearance is a good thing.

One interesting issue is when you have rolling element main bearings, which thankfully is not the case for most small aircraft engines. When the aluminum crankcase changes size due to temperature changes, it is possible for the crankshaft to be pinched lengthwise, or even come loose lengthwise between the main bearings. For this reason when you assemble one of these engines you may shim the crankshaft to a negative thrust clearance, i.e. pinched a little at room temperature.

It is the opposite : In “modern” car engines or truck diesels the pistons are designed to have minimal variation if cold or hot. To this purpose steel elements are cast into the piston to keep it collapsing when cold. So the steel is in compression while the piston is cold. When warming up first the stress in the alu piston reduces before it expands to operating temperatures. That is why pistons with this design require less undersize in the bore. Yes, it is a shame that no aero engine has alu cylinders, even maybe Nikasil coated like BMW bikes. Cast iron liners in alu would be quite good too, and the whole affair should be clamped with long through bolts and nuts on the cylinder head to save the cylinders from being stressed. This would have pistons running with a lot less variations in clearance. Roller or ball bearings for cranks should be selected with one bearing the positioning type and the corresponding either a loose fit or a roller type, so that would be a bad design if temperature was critical. Rotax dimensions are small but then the amount of temp. varaition is small as well. No problem. Vic

Prior to seeing the aluminum Lycoming cylinders, I did some research into (diesel style) steel pistons that would match the thermal expansion of the steel cylinders. On a boxer engine it seemed like an interesting idea because of the 100% primary balance, but the con rod inertial forces would be high. Higher revving engines use aluminum pistons, in Nikasil or similar, like the air cooled BMW motorcycle engines Vic mentions.

A lot of rolling element main bearing engines are temperature critical in terms of thrust clearance regardless of it being ideal design practice (or not!). I was surprised when I first saw it on a Ducati V-twin.

Don´t mention bevel gear Ducatis. They are beauty but in Italy – and elsewhere – famous for standing only very modest mileage till breakdown in various places. I´d be able to tell stories from last millennium ….. You will have seen shouldered ball bearings for mains I guess. They should have applied big standard ball bearings and rollers the other side. There is some axial load from the primary gears, so one ball bearing is there required. As to main bearings in aluminium cases: One big issue is the increase of clearance in the plain bearings when the case gets up to working temperatures. For a reason car engines have cast iron crankcases to keep the clearances constant to avoid oil pressure drop when hot. So when you use alu for the case you should better cast in bearing holders made in steel or cast iron so temperatures don´t effect bearing clearances in a range from 20 degrees up to maybe 100 degrees. I do not know how Lyco etc. deal with this, haven´t seen an open engine yet. Vic

Vic, I’m impressed that you guessed bevel gear Ducatis… Those are the ones that specify a slightly negative clearance at room temperature for crankshaft end float. The later non-bevel one-piece crankshaft engines engines use a similar main bearing setup but without the negative clearance. I got about 60,000 km out of my first bevel Ducati prior to overhaul, in the mid-90s… But I did learn a lot doing the built up crankshaft myself.

Oil film bearing design for thermally varying clearnace sounds like an interesting problem to me. There’s a lot to this machinery design stuff, eh?