Housing material

[JeanMicheling]

Here another question (I know, I'm asking a lot of questions!). What would be the best spindle housing material. I know that cast iron and malleable cast iron are used a few. Cast iron is a little bit more thermically stable than malleable. Is it a good thing? Personnaly, I would say yes because thermal expension is lowered and the less it moves, the better it is. What's your opinion?

 

[EnglishMuffin]

If possible, I would use steel, although cast iron is easier to machine and cheaper in quantity, especially if the shape required is complex. Differences in thermal properties are probably relatively insignificant compared to other variables.

 

[Mendit]

Ductile iron works great especialy if there is room for thermal growth. When the bearing at 1 end is allowed to move, or rotate. The semi-self lubrication properties of iron are much better than steel. There is also consideration of steel hardness. If your housing is too hard you will get fatigue set up between the 2 steels which causes failures.

 

[JeanMicheling]

Mendit, what do you mean by semi-self lubrication properties pf iron and steel? Do you mean the capacity of the material to allow the bearing outer ring to slip on the housing bore?

And what about a spindle that run at colder temp. Some of our clients have their machine in a cold environnement (about 0 celcius). The spindle is shutted down for the weekend, stay around 0 celcius. When they turn the motor on, the spindle reach 3600 rpm in less than a few seconds so I wonder if, because the bearing outer ring expands faster than the housing bore (which is in ductile iron), it will shrink the bearing. Furthermore, the spindle is at once loaded before stabilizing its temp. Can those phenomenas be an issue?

 

[Tmoose]

I think you/they need to keep those spindles warm.

What size are the bearings?

If the cold housing is thick around the bearing, it will squeeze the life out of the bearing like a boa constrictor

 

[Mendit]

Are you working out side ? are your employees working in temperatures of less than 16C. Running high speed machinery with great preciison requires the temperatures to be even right across the operation temperature range.
If the spindles are cold have a heated oil system, if the spindles get too hot use the oil system to keep them temperature stable. Most precision grinder have temperature controlled systems

The Graphite content of ductile iron reduces the coefficient of friction between the bearing O/d and housing I/d

 

[JeanMicheling]

My spindle is running at 3600 RPM and usually the temp. is controlled. But there are some places where the machine runs at cold temp. in the winter. I don't need too much precision cause I'm just shooting woods. I'm more concerned about the impact that it could have on my spindle reliability. I guess that a failure caused by this phenomena would show a premature lubricant deterioration leading to some abrasive wears?

EnglishMuffin, it seems that ductile iron works better than steel so why you'd rather use steel?

 

[Hup]

Does anyone have some experience with housings equipped with brass sleeve inside? I've read some interesting papers refering this solution quite advantageous in applications where solid preload of bearing set is used. Due to larger thermal expansion it should mitigate the "boa effect" on the outer ring.

But there may be some drawbacks not mentioned in the papers, and the housing cost.

 

[Tmoose]

Are those papers available online? I'd like to read them.

Greater CTE >could< be helpful in a gradually increasing temperature situation with a very thin housing. The European solution is to warm spindles up slowly. American machine shops expect their spindles to go from 0 to 60 in 5 seconds.

I think the generally greater thermal expansion of brass and bronze would be a dis-advantage on a high speed spindle if the housing is "thick walled." I believe "Thick walls" are a problem for high speed spindles with iron and steel housings. "Thick walled" would have to be defined using a series of FEA analyses and experiments. At a minimum a housing thickness should be 3 or 4 times the bearing outer race thickness. The RAPIDLY warming bearing passes some thermal energy to the housing. The heat conduction will lag behind the temp rise (*), and the housing inner bore, near the bearing will be warmer than the outer sections. For a while at least the warm section is thin compared to the cool section, and thus will have to expand inward, reducing bore diameter, increasing preload and temp generation.

A "high speed" preloaded spindle bearing gets warm rapidly, before the housing temp can rise at all, and can cascade into thermal run-away smoking junk in just a few minutes.

(*)Add large amounts of heat energy by Playing a torch on one end of a 12 inch bar of iron, bronze or even aluminum while holding the other end in your bare hand. It is possible to melt one end before the other end becomes uncomfortably hot

--
Dan Timberlake

 

[JeanMicheling]

When I look at some spindles like Gilman's for exemple, they got thick walled cartridge spindle and they sell it with a square housing (where the cartridge spindle goes in) and you don't have more than 0.00025'' clearance between cartridge outer diameter and housing bore. they are high speed spindles and greased. I don't think the housing thikness is so important as far as they keep it at room temperature. I'd say it is more problematic in the kind of application I have because spindles are sometime kept cold and and they start loading it without letting it warming up. Don't you think it would be better to use a more conductive material as aluminium. Thus the inward expansion will be minimized since the steady state will show up faster. But in this case, using a housing to hold the cartridge spindle with a slight clearance might be wrong. I'd rather make a spindle with a flange at about the center of mass so I won't need to guide the cartridge with a fit.

 

[Hup]

The paper I mentioned is from Jedrzejewski.

It seems to me, that the brass sleeve takes the advatage of CTE together with better thermal conductivity. When the sleeve has larger axial length it coud provide a non-constraining heat sink for the outer rings.

One of conclusions in the paper could be interesting for JaenMicheling:
it is probably not good idea to make spindle housings of polymer-concrete. :-/