gearboxes
#1
Posted 24 September 2012 - 08:14
We have also some in WSbR because of the engine is too powerful for the gearbox. But what's the matter in F1?
Advertisement
#2
Posted 24 September 2012 - 08:27
We have a lot of gearbox problems in F1 this season. Why? Why are gearboxes so complicated.
We have also some in WSbR because of the engine is too powerful for the gearbox. But what's the matter in F1?
To stop the drivers breaking them...
#3
Posted 24 September 2012 - 09:15
#4
Posted 24 September 2012 - 09:40
There are electronic controls, hydraulics, pumps and gears, shafts and seals, they undoubtedly stress their lubricants to the greatest possible degree.
Are they still running carbon fibre casings?
#5
Posted 24 September 2012 - 11:29
when the suspension arms are typically mounted on the gearbox-casing.
#6
Posted 24 September 2012 - 13:02
#7
Posted 24 September 2012 - 13:23
Edited by Rasputin, 24 September 2012 - 13:31.
#8
Posted 24 September 2012 - 19:15
#9
Posted 24 September 2012 - 20:05
If that is the case then some very basic E.M.C. (electro magnetic compatibility) mistakes have been made.Power cables buried under the track can cause havoc with the Moog valves controlling the gear change actuator if they are not sufficiently shielded.
#10
Posted 24 September 2012 - 20:09
#11
Posted 24 September 2012 - 22:10
#12
Posted 24 September 2012 - 22:44
Edited by gruntguru, 24 September 2012 - 22:45.
#13
Posted 24 September 2012 - 22:58
I remember an undergraduate mechanical design class where we all designed an industrial gearbox. Several students including myself were surprised at how large the gearbox needed to be, compared to automotive transmissions designed to handle higher power levels. The answer lies in the very short time any particular ratio spends handling maximum power or torque, whereas an industrial gearbox must cope with its rated power 24/7.I'm surprised - nearly said astonished - that gearboxes, especially F1 gearboxes, don't spend all their time exploding. When you consider what they have to do, and how small they are, I'm impressed.
As for F1 gearboxes, the secret to the small size is maximising speeds thus minimising torques and forces within the box.
#14
Posted 25 September 2012 - 01:36
I'm surprised - nearly said astonished - that gearboxes, especially F1 gearboxes, don't spend all their time exploding. When you consider what they have to do, and how small they are, I'm impressed. I don't know the statistics, but I'm sure a good deal of the failures are due to hydraulics as much as mechanical failure.
The primary mechanical components in the gearbox (gears, shafts, bearings, dogs, etc.) are very reliable. Instead, I'd propose that it's the sensors, electrohydraulic devices, software, etc. that contribute to failures. The typical differential alone is now probably more complex than the rest of the gearbox.
#15
Posted 25 September 2012 - 06:00
I know. It doesn't stop me being surprised. Perhaps I'm just easily surprised.The primary mechanical components in the gearbox (gears, shafts, bearings, dogs, etc.) are very reliable.
#16
Posted 25 September 2012 - 06:16
The Rpm is high of course, 18000 Rpm, but that does not affect the lifespan of gears and bearing nearly as much as the torque.
Roller bearings, life has an xponential relation to load, which comes from the torque, like; L10h = (C/P)^3.333 * 10^6/(Rpm * 60).
Where L10h is the nominal lifespan in hours, C is the bearing's dynamic load rating and P is the equivalent load.
The relation is similar for fatigue of the gears, while both are actually building a better oil-film the higher the speed.
#17
Posted 25 September 2012 - 06:52
#18
Posted 25 September 2012 - 07:49
Teams rarely quote a gearbox-failure as anything but a "gearbox-failure", I've never heard any detailed xplanation either,I don`t think i have ever heard about a bearing failure in F1.
but typically it's gears and bearings that would be the topic for lifespan optimizations in order to save space and weight?
Wheel-bearings fail now and again however.
Edited by Rasputin, 25 September 2012 - 14:23.
#19
Posted 25 September 2012 - 14:04
No it's the seamless shifting.
That's just a marketing term!
Advertisement
#20
Posted 28 September 2012 - 00:29
One should remember the relatively low input-torque involved in an F1-gearbox, no more than some 300 Nm (220 lb.ft) actually.
The Rpm is high of course, 18000 Rpm, but that does not affect the lifespan of gears and bearing nearly as much as the torque.
Roller bearings, life has an xponential relation to load, which comes from the torque, like; L10h = (C/P)^3.333 * 10^6/(Rpm * 60).
Where L10h is the nominal lifespan in hours, C is the bearing's dynamic load rating and P is the equivalent load.
The relation is similar for fatigue of the gears, while both are actually building a better oil-film the higher the speed.
Rasputin,
With the hydrodynamic fluid film contact in gear meshes, one must also consider the power being transferred and the efficiency of the process. With a gear mesh that is transferring large amounts of power at high speed and low torque, there may be enough flash temp rise within the hydrodynamic oil film to result in loss of viscosity and significant thinning of the dynamic oil film thickness, thus resulting in scoring failure of the gears.
slider
#21
Posted 28 September 2012 - 03:45
but the lower viscosity will in itself lead to lower losses again due to less shear-forces in the oil-film, why a balance will be found eventually.
Needless to say, this is why F1 cars have a cooling-circuit for the gearbox oil, not so sure if it's separate or integrated with the engine's?
#22
Posted 28 September 2012 - 07:57
Teams rarely quote a gearbox-failure as anything but a "gearbox-failure", I've never heard any detailed xplanation either,
but typically it's gears and bearings that would be the topic for lifespan optimizations in order to save space and weight?
Wheel-bearings fail now and again however.
Gearbox bearings i doubt. apart from the diff during a bump. ofc they push their limits but bearings is a serious science that make them carbon wishbones look like lego.
Wheelbearings i doubt go off without a bump.
Never heard about them failing either. Suspension goes first it seems.
#23
Posted 29 September 2012 - 05:12
This is where rules saying normal alloys for all major components, engines, transaxles and suspension components would save a great deal of money and increase reliuability quite a lot. And maybe a 25kilo penalty!
Then get rid of all the electronic gadgetry associated with transmissions and even more reliability and even les cost. And very possibly better racing and more emphasis on the drivers skill.
#24
Posted 29 September 2012 - 07:35
a failed dynamic contact-seal could lead to an almost instant failure. Wasn't it Mika Hakkinen who's McLaren had a spectacular wheel-bearing failure in the early 00s?
#25
Posted 29 September 2012 - 20:03
I suspect the transmission woes is the age old thing of building the lightest components they think they can get away with. And seemingly they do most of the time. As others have said banging wheels against the scenery, stationary or moving will be most of the problems encountered. Carbon fibre and very light alloys would never be my choice for a gearbox case as it does not like being banged around, and especially with the suspension mounted to it. But if one has it they all must.
This is where rules saying normal alloys for all major components, engines, transaxles and suspension components would save a great deal of money and increase reliuability quite a lot. And maybe a 25kilo penalty!
Then get rid of all the electronic gadgetry associated with transmissions and even more reliability and even les cost. And very possibly better racing and more emphasis on the drivers skill.
There is actually for F1 a reasonable degree of differentiation between the teams in terms of material choice for the gearbox case;
Carbon fibre with metallic bulkheads - Should be stiffest, lightest and most robust and can be repaired relatively easily but expensive to produce in house and has moderate lead time.
Cast Titanium - Strong but difficult to keep the weight down without etching also difficult to repair and has long lead time.
Cast Aluminium - Weak with poor fatigue strength but cheap and fairly short lead time.
Bearings rarely fail catastrophically but the races can become damaged by interruption to the oil supply and then the damaged races allow the gear shafts to move excessively which can cause the gear teeth to fail. Teams try to run the absolute minimum amount of oil in the gearbox circuit - firstly because of it's weight but secondly because churning the oil increases it's temperature and causes a power loss.
#26
Posted 29 September 2012 - 20:18
#27
Posted 29 September 2012 - 23:11
If a radial bearing's races become damaged to the extent that the shafts gets out of line, that bearing is already gone off to tribology-heaven, that much is for certain.
As Rachael noted, rolling element bearings usually do not fail catastrophically. The most common failure mode is a race surface spall initiated by a subsurface shear fracture. This is a classic rolling element bearing fatigue failure and it is usually a rather benign event. The spalling produces tiny bits of ferrous debris which can easily be monitored for using a simple magnetic chip detector. The chip detector will indicate a spalling failure long before it becomes severe enough to cause shaft dislocation.
The gearbox internal rotating components (bearings, gears, etc.) are mostly oil-cooled, since the conductive heat path away from these parts is quite limited. Other internal components such as the outer bearing races can conduct heat out thru the housing. However, if the housing is made from composite or a metal with low thermal conductivity like titanium as opposed to aluminum, then the heat rejection issue becomes more difficult. Conventional bearing alloy steels (like E52100) and carburizing gear steel alloys (like 9310) are usually heat treat stabilized for temperature limits of around 350 to 400 degF. So a loss of lubricant flow and cooling can easily result in temperatures high enough to degrade the mechanical properties of gear teeth and bearing races.
#28
Posted 30 September 2012 - 05:11
As Rachael noted, rolling element bearings usually do not fail catastrophically. The most common failure mode is a race surface spall initiated by a subsurface shear fracture. This is a classic rolling element bearing fatigue failure and it is usually a rather benign event. The spalling produces tiny bits of ferrous debris which can easily be monitored for using a simple magnetic chip detector. The chip detector will indicate a spalling failure long before it becomes severe enough to cause shaft dislocation.
The gearbox internal rotating components (bearings, gears, etc.) are mostly oil-cooled, since the conductive heat path away from these parts is quite limited. Other internal components such as the outer bearing races can conduct heat out thru the housing. However, if the housing is made from composite or a metal with low thermal conductivity like titanium as opposed to aluminum, then the heat rejection issue becomes more difficult. Conventional bearing alloy steels (like E52100) and carburizing gear steel alloys (like 9310) are usually heat treat stabilized for temperature limits of around 350 to 400 degF. So a loss of lubricant flow and cooling can easily result in temperatures high enough to degrade the mechanical properties of gear teeth and bearing races.
I believe that chip-detecting is a bit too-late measure, there are more sophisticated pro-active condition-monitoring methods today, like SKF's Aptitude Asset;
http://www.skf.com/p...t...076&lang=en
Oil-cooling is obviously the way to convey the losses, material hysteresis and viscous loss, even if you figure that an F1 gearbox has only a one percent loss, it's still 5+ kW, like ten toasters, to remove.
Edited by Rasputin, 30 September 2012 - 11:38.
#29
Posted 30 September 2012 - 10:37
There is actually for F1 a reasonable degree of differentiation between the teams in terms of material choice for the gearbox case;
Carbon fibre with metallic bulkheads - Should be stiffest, lightest and most robust and can be repaired relatively easily but expensive to produce in house and has moderate lead time.
Cast Titanium - Strong but difficult to keep the weight down without etching also difficult to repair and has long lead time.
Cast Aluminium - Weak with poor fatigue strength but cheap and fairly short lead time.
Bearings rarely fail catastrophically but the races can become damaged by interruption to the oil supply and then the damaged races allow the gear shafts to move excessively which can cause the gear teeth to fail. Teams try to run the absolute minimum amount of oil in the gearbox circuit - firstly because of it's weight but secondly because churning the oil increases it's temperature and causes a power loss.
Rach - "its", not "it's".
#30
Posted 30 September 2012 - 12:56
#31
Posted 30 September 2012 - 13:48
Yeah, right. When someone makes a fundamental error like that you have to question what they know about gearboxes.
Yes - exactly - couldn't agree more.
#32
Posted 30 September 2012 - 14:34
#33
Posted 30 September 2012 - 14:36
I'd hope on a technical forum that my posts are read for the technical content rather than use of English - perhaps the former is over your head hence why you only comment on the latter.Rach - "its", not "it's".
I admit to being irritated by poor use of grammar by others, but would not be so rude as to bring attention to it in public.
#35
Posted 30 September 2012 - 15:40
I see what you did thereGeez, if we only had some sort of gearbox expert around here to answer ... I'm sure this is related to that ancient layshaft technology!
#36
Posted 30 September 2012 - 17:53
Anyway, the problem with Aluminium is not really the strength or the weight, it's actually 40% lighter that Ti, but the stiffness and most of all thermal expansion.
Thermal expansion of Alu is three times Ti and does not match CFRP, why they cannot be molded together where temperature will be high, such as in a gearbox.
Ti and CFRP works great together however, why those CFRP-gearboxes always have Ti-inserts for bearings and such.
#37
Posted 30 September 2012 - 22:24
Cast Aluminium - Weak with poor fatigue strength but cheap and fairly short lead time."
Is the difference in lead time the difference in tools they need because they gravity cast the aluminium and low pressure diecast the Titanium?
#38
Posted 01 October 2012 - 00:23
KC, it's "Rach - it's "its", not "it's".".Rach - "its", not "it's".
Tsk - standards are dropping 'round here.
#39
Posted 01 October 2012 - 04:04
I'd hope on a technical forum that my posts are read for the technical content rather than use of English - perhaps the former is over your head hence why you only comment on the latter.
I admit to being irritated by poor use of grammar by others, but would not be so rude as to bring attention to it in public.
Probably both over my head and not of any particular interest to me. So I commented on the only bit that was of interest to me.
I have discovered over many years of experience that, regrettably, sheilas have little or no sense of humour. (If Rach is indeed a sheila).
Advertisement
#40
Posted 01 October 2012 - 04:06
KC, it's "Rach - it's "its", not "it's".".
Tsk - standards are dropping 'round here.
You are right - that's what I should have written - much cleverer.
#41
Posted 01 October 2012 - 05:22
An Aluminium gearbox casing would most certainly be pressure diecast, like wheel-rims, but I believe that the melt in itself is more complicated with Ti, there are far more Al-foundries around anyway.Rachael"Cast Titanium - Strong but difficult to keep the weight down without etching also difficult to repair and has long lead time.
Cast Aluminium - Weak with poor fatigue strength but cheap and fairly short lead time."
Is the difference in lead time the difference in tools they need because they gravity cast the aluminium and low pressure diecast the Titanium?
#42
Posted 01 October 2012 - 19:10
I have discovered over many years of experience that, regrettably, sheilas have little or no sense of humour. (If Rach is indeed a sheila).
I can well imagine that 'sheilas' in your company would appear to have no sense of humour - the fault would be yours not theirs however. What have you been crossed with?
#43
Posted 01 October 2012 - 19:19
Rachael"Cast Titanium - Strong but difficult to keep the weight down without etching also difficult to repair and has long lead time.
Cast Aluminium - Weak with poor fatigue strength but cheap and fairly short lead time."
Is the difference in lead time the difference in tools they need because they gravity cast the aluminium and low pressure diecast the Titanium?
Greg it could be the difference in tools, the Aluminium box I am aware of was gravity cast but I've not been involved with a cast Ti one. I believe the design release for a Ti casting is something over 2 months earlier than an Ally casting which means you have to commit to your rear suspension geometry earlier.
The next 'big thing' may be DMLS - Direct Metal Laser Sintering which allows partial or fully voided structures that could never be manufactured by machining or casting. At the moment it's limited to smaller parts and I think the fatigue strength is poor due to rough surface finish.
Edited by rachael, 01 October 2012 - 19:25.
#44
Posted 01 October 2012 - 23:30
That does sound like the time difference between gravity cast and LP diecast tooling.You could X ray all the aluminium ones and throw the ones with voids away and still be ahead cost wise.Greg it could be the difference in tools, the Aluminium box I am aware of was gravity cast but I've not been involved with a cast Ti one. I believe the design release for a Ti casting is something over 2 months earlier than an Ally casting which means you have to commit to your rear suspension geometry earlier.
The next 'big thing' may be DMLS - Direct Metal Laser Sintering which allows partial or fully voided structures that could never be manufactured by machining or casting. At the moment it's limited to smaller parts and I think the fatigue strength is poor due to rough surface finish.
I guess if they can sinter conrods then other components can't be far away.
Edited by Greg Locock, 02 October 2012 - 00:23.
#45
Posted 02 October 2012 - 03:04
What have you been crossed with?
God.
#46
Posted 02 October 2012 - 03:09
The Ferrari F2000 F1 gearbox was fabricated from numerous titanium parts, but strengthened by a large carbon moulding that bolted on top and provided suspension pick-up points and housed the torsion bars.
Tony Matthews,
Thanks for the nice illustration of the 2000 Ferrari gearbox housing pieces. Obviously, as you mentioned, the structure was welded. Do you happen to know what welding process was used (EB, laser, TIG, etc.)? Welding high strength titanium alloys is quite difficult due to high reactivity with oxygen when molten. TIG welding is the most economical method, but even TIG requires the titanium welding to be performed in an enclosure purged with inert gas.
Investment casting can be used to produce a high precision, lightweight titanium housing. The wax masters can be produced fairly rapidly using stereolithography directly from the digital model. The housing walls and ribs can be as thin as 1.5mm with titanium.
slider
#47
Posted 02 October 2012 - 04:14
Tony Matthews,
TIG welding is the most economical method, but even TIG requires the titanium welding to be performed in an enclosure purged with inert gas.
slider
I think Ti alloy can be welded using TIG equipment without using an Argon filled enclosure. Bicycle frames are routinely welded up that way. The days of guys in breathing apparatus welding Ti in O2 free rooms or glove boxes are relics of the '50 and '60s. You can see a guy welding one up this way in this video: The interior of the tubes are purged of air and fed pressurized Argon through plugs in the tube ends.
#48
Posted 02 October 2012 - 06:55
The Ferrari box was made from a mix of investment castings and CNC macined bits, and was welded in an inert gas chamber. I didn't put that into my original post as I try to be economical. Some of the parts are amazingly thin, and the welds were beautiful.Tony Matthews,
Thanks for the nice illustration of the 2000 Ferrari gearbox housing pieces. Obviously, as you mentioned, the structure was welded. Do you happen to know what welding process was used (EB, laser, TIG, etc.)? Welding high strength titanium alloys is quite difficult due to high reactivity with oxygen when molten. TIG welding is the most economical method, but even TIG requires the titanium welding to be performed in an enclosure purged with inert gas.
Investment casting can be used to produce a high precision, lightweight titanium housing. The wax masters can be produced fairly rapidly using stereolithography directly from the digital model. The housing walls and ribs can be as thin as 1.5mm with titanium.
slider
#49
Posted 02 October 2012 - 19:58
That does sound like the time difference between gravity cast and LP diecast tooling.You could X ray all the aluminium ones and throw the ones with voids away and still be ahead cost wise.
I guess if they can sinter conrods then other components can't be far away.
I think sintered conrods is a different process using a die, heat and pressure. DMLS needs no die and is like 3D printing.
#50
Posted 04 October 2012 - 19:55