Quote from: Wolfram66 on 12/29/2019 04:01 pmQuote from: livingjw on 12/29/2019 03:06 pmQuote from: codav on 12/29/2019 04:36 amQuote from: su27k on 12/29/2019 03:11 amThe following is cropped from bocachicagal's photos, seems that each tile is supported by 3 bolts, corresponding to the 3 holes on the tile.I'm quite sure this way of mounting the heat shield tiles won't be the final method. While it works and allows for quickly attaching and detaching the tiles, the holes actually create weak spots in the tile, will probably add tension due to different thermal expansion coefficients of the tiles and the underlying steel structure and also trap some of the super heated plasma inside the cavity, probably melting the bolt.I disagree. To me this seems a good approach.- The tiles have an outer covering for strength and toughness. Inside is insulation. - Without knowing how the tile shell is designed, you can't say the holes will cause weak spots. They didn't just drill holes in the tile shell. I am sure they are designed to handle the loads.- Thermal expansion was surely taken into consideration when engineering the fasteners and attachment. Why would they not?- Attachment points are recessed and can be easily plugged with insulating material.JohnJohn,If you inspect the left most tile from image above, you can see that it has fractured between the 3 bolt holes used to affix the tile.That feature looks too uniform to have just "fractured", it looks intentional.
Quote from: livingjw on 12/29/2019 03:06 pmQuote from: codav on 12/29/2019 04:36 amQuote from: su27k on 12/29/2019 03:11 amThe following is cropped from bocachicagal's photos, seems that each tile is supported by 3 bolts, corresponding to the 3 holes on the tile.I'm quite sure this way of mounting the heat shield tiles won't be the final method. While it works and allows for quickly attaching and detaching the tiles, the holes actually create weak spots in the tile, will probably add tension due to different thermal expansion coefficients of the tiles and the underlying steel structure and also trap some of the super heated plasma inside the cavity, probably melting the bolt.I disagree. To me this seems a good approach.- The tiles have an outer covering for strength and toughness. Inside is insulation. - Without knowing how the tile shell is designed, you can't say the holes will cause weak spots. They didn't just drill holes in the tile shell. I am sure they are designed to handle the loads.- Thermal expansion was surely taken into consideration when engineering the fasteners and attachment. Why would they not?- Attachment points are recessed and can be easily plugged with insulating material.JohnJohn,If you inspect the left most tile from image above, you can see that it has fractured between the 3 bolt holes used to affix the tile.
Quote from: codav on 12/29/2019 04:36 amQuote from: su27k on 12/29/2019 03:11 amThe following is cropped from bocachicagal's photos, seems that each tile is supported by 3 bolts, corresponding to the 3 holes on the tile.I'm quite sure this way of mounting the heat shield tiles won't be the final method. While it works and allows for quickly attaching and detaching the tiles, the holes actually create weak spots in the tile, will probably add tension due to different thermal expansion coefficients of the tiles and the underlying steel structure and also trap some of the super heated plasma inside the cavity, probably melting the bolt.I disagree. To me this seems a good approach.- The tiles have an outer covering for strength and toughness. Inside is insulation. - Without knowing how the tile shell is designed, you can't say the holes will cause weak spots. They didn't just drill holes in the tile shell. I am sure they are designed to handle the loads.- Thermal expansion was surely taken into consideration when engineering the fasteners and attachment. Why would they not?- Attachment points are recessed and can be easily plugged with insulating material.John
Quote from: su27k on 12/29/2019 03:11 amThe following is cropped from bocachicagal's photos, seems that each tile is supported by 3 bolts, corresponding to the 3 holes on the tile.I'm quite sure this way of mounting the heat shield tiles won't be the final method. While it works and allows for quickly attaching and detaching the tiles, the holes actually create weak spots in the tile, will probably add tension due to different thermal expansion coefficients of the tiles and the underlying steel structure and also trap some of the super heated plasma inside the cavity, probably melting the bolt.
The following is cropped from bocachicagal's photos, seems that each tile is supported by 3 bolts, corresponding to the 3 holes on the tile.
Given that SpX is known for 3-D printing, they might try printing the ceramic tiles and bolts as one whole piece.
The purpose would be to avoid different expansion coefficients between bolt and tile. Nuts could also be of same material. The bolt holes in the underlying steel sheet would be sized to accommodate greater expansion during heating.Added: It is OK that holes will expand larger during heating, but the holes will also expand further apart, which could exert excessive force on the bolts.
Given that SpX is known for 3-D printing, they might try printing the ceramic tiles and bolts as one whole piece. The purpose would be to avoid different expansion coefficients between bolt and tile. Nuts could also be of same material. The bolt holes in the underlying steel sheet would be sized to accommodate greater expansion during heating.Added: It is OK that holes will expand larger during heating, but the holes will also expand further apart, which could exert excessive force on the bolts.
[...]twitter.com/erdayastronaut/status/1211550091414265856Quote It’d be so cool if Tesla motors were powerful enough to spin start! Have you moved onto direct drive / electromechanical on the body flaps with motors yet or still spinning a pump for SN1?https://twitter.com/elonmusk/status/1211556824442425344Quote Direct drive using several Tesla Plaid motors in parallel for SN1. Simpler, lighter & more fault tolerant. Rear flaps each need ~1.5 megawatts. It’s like moving the entire wing of an aircraft!
It’d be so cool if Tesla motors were powerful enough to spin start! Have you moved onto direct drive / electromechanical on the body flaps with motors yet or still spinning a pump for SN1?
Direct drive using several Tesla Plaid motors in parallel for SN1. Simpler, lighter & more fault tolerant. Rear flaps each need ~1.5 megawatts. It’s like moving the entire wing of an aircraft!
How about worm gears for the rear flaps like the jack screws in planes with an upper rear elevator on their tails?Could you make them strong enough because getting enough torque to drive them at such a high gear ratio wouldn't be a issue.
Many ways to solve this problem, but the power requirements are much higher than aircraft control surfaces. When moving giant body flaps rapidly, achieving high power is much harder than high torque.
So we’re gonna see the plaid power train in action in a spacecraft before we see it in a production tesla? 🤣😂🤣
Building prototypes is relatively easy, volume production is hard
Continuing discussion on driving Starship flapsQuote from: FutureSpaceTourist on 12/30/2019 07:05 am[...]twitter.com/erdayastronaut/status/1211550091414265856Quote It’d be so cool if Tesla motors were powerful enough to spin start! Have you moved onto direct drive / electromechanical on the body flaps with motors yet or still spinning a pump for SN1?https://twitter.com/elonmusk/status/1211556824442425344Quote Direct drive using several Tesla Plaid motors in parallel for SN1. Simpler, lighter & more fault tolerant. Rear flaps each need ~1.5 megawatts. It’s like moving the entire wing of an aircraft!twitter.com/john_gardi/status/1211567070808346624Quote How about worm gears for the rear flaps like the jack screws in planes with an upper rear elevator on their tails?Could you make them strong enough because getting enough torque to drive them at such a high gear ratio wouldn't be a issue.https://twitter.com/elonmusk/status/1211784060705656832Quote Many ways to solve this problem, but the power requirements are much higher than aircraft control surfaces. When moving giant body flaps rapidly, achieving high power is much harder than high torque.twitter.com/shaneappleton7/status/1211557213610733568Quote So we’re gonna see the plaid power train in action in a spacecraft before we see it in a production tesla? 🤣😂🤣https://twitter.com/elonmusk/status/1211780010664742923Quote Building prototypes is relatively easy, volume production is hard
QuoteDirect drive using several Tesla Plaid motors in parallel for SN1. Simpler, lighter & more fault tolerant. Rear flaps each need ~1.5 megawatts. It’s like moving the entire wing of an aircraft!This tells me the motor case or it's shaft will be attached to the flap, and the other attached to the body. The motor won't even be able to make a half revolution. On the other hand it will be able to move that flap fast.
Quote from: Eka on 12/31/2019 03:04 amQuoteDirect drive using several Tesla Plaid motors in parallel for SN1. Simpler, lighter & more fault tolerant. Rear flaps each need ~1.5 megawatts. It’s like moving the entire wing of an aircraft!This tells me the motor case or it's shaft will be attached to the flap, and the other attached to the body. The motor won't even be able to make a half revolution. On the other hand it will be able to move that flap fast.Can those motors produce the kN-m range torques needed to twist the flap against airflow? I think they will need extra gear reduction drives. Musk mentioned worm gears at the Mk.1 event, IIRC.
Quote from: envy887 on 12/31/2019 01:55 pmQuote from: Eka on 12/31/2019 03:04 amQuoteDirect drive using several Tesla Plaid motors in parallel for SN1. Simpler, lighter & more fault tolerant. Rear flaps each need ~1.5 megawatts. It’s like moving the entire wing of an aircraft!This tells me the motor case or it's shaft will be attached to the flap, and the other attached to the body. The motor won't even be able to make a half revolution. On the other hand it will be able to move that flap fast.Can those motors produce the kN-m range torques needed to twist the flap against airflow? I think they will need extra gear reduction drives. Musk mentioned worm gears at the Mk.1 event, IIRC.Wheel torque in the acceleration range they're targeting for the reminder plaid models is 10-15 kNm per wheel.
Quote from: niwax on 12/31/2019 02:16 pmQuote from: envy887 on 12/31/2019 01:55 pmQuote from: Eka on 12/31/2019 03:04 amQuoteDirect drive using several Tesla Plaid motors in parallel for SN1. Simpler, lighter & more fault tolerant. Rear flaps each need ~1.5 megawatts. It’s like moving the entire wing of an aircraft!This tells me the motor case or it's shaft will be attached to the flap, and the other attached to the body. The motor won't even be able to make a half revolution. On the other hand it will be able to move that flap fast.Can those motors produce the kN-m range torques needed to twist the flap against airflow? I think they will need extra gear reduction drives. Musk mentioned worm gears at the Mk.1 event, IIRC.Wheel torque in the acceleration range they're targeting for the reminder plaid models is 10-15 kNm per wheel.Elon specifies explicitly "direct drive", meaning no gear. Model 3 motor has peak power 250 kW and peak torque 450 Nm:https://en.wikipedia.org/wiki/Tesla_Model_3Powerwise, they need 6 motors. Unfortunately, my "back of envelope" implies torque at the MNm range. Either from Elon's 1.5 MW and assuming angular velocity in the order 1/sec. Or, from assuming 5g decelleration of a 100 metric ton spacecraft, which implyies 5 MN decelleration force. If one flap carries 1 MN, and we assume a moment arm in the order of 1 m, then we arrive to MNm, again. Which is in clear contrast to the cc. 3 kNm torque of 6 Model 3 motors. So, I don't know.
[snip]His comment regarding direct drive was in response to a question about hydraulic actuators. I would expect them to have at least their standard integrated gear box.
Quote from: niwax on 12/31/2019 03:29 pm[snip]His comment regarding direct drive was in response to a question about hydraulic actuators. I would expect them to have at least their standard integrated gear box.There will have to be gearing if for no other reason to enable accurate & repeatable flap movements. X shaft revolutions per degree of flap movement. Can't get that kind of control from an electric motor over fractions of a revolution.Not to mention the added benefit of trading motor revolutions for increased force.[edit - grammar]
I work for a company that is a leader in direct drive motors. By definition it means that the torque is directly coupled to the load - no gearing, pulleys or shafts. Servo motors can maintain fraction of arc second accuracy, much less degrees. That is not a problem. Whether they can actually get sufficient torque in this manner is a very interesting question.
SpaceX will have the budget for the highest accuracy encoders and drives, but from a motor control standpoint, control is much, much easier if there is some gear reduction between the motor and the load.On a car or a washing machine, you don't need stepper motor behavior, don't need to control wheel rotation accurately to degrees to drive the car. Direct drive can work fine.But on a control surface, position must be accurate, and aero forces on the flaps pushing back against the motor can create a lot of encoder noise. When encoder values are changing the opposite direction of what the drive is commanding, the drive either has to shut down or the drive has to be able to interpret and ride through the disturbance. Even a 10:1 gearbox makes a big difference on reducing backlash to the encoder.I would not be surprised if they end up with some gearing.