I am leaning away from the ball screws due to the potential for rust and corrosion over the life of the system. It needs to be maintenance free for 10+ years. Finding and machining anti corrosion ball screws and ball nuts in the lengths I need is very costly. I am leaning towards a complete chain drive system with a worm gear off the motor gearbox, which prevents the load from moving with gravity in a power outage and eliminates the need for braking. Of course if a chain breaks that is a different story. All good ideas above so thanks guys.
Actually, ballscrews have much lower friction than worm gears or acme screws, although higher friction than an equivalent geared transmission. Ball screws were used on gamma cameras to lift heads that could weigh over 3,000 pounds, and in tensile testers up to about 2000 pounds.
Sorry, my bad, you are right, I was wrong. I was thinking of leadscrews, which are dreadful things with sliding friction. Ballscrews have much more efficient rolling contact, as used in the recirculating ball steering box in my '67 Corvair!
My favorite worm reducer is the Cone Drive but they can be a bit spendy.
Yes, they are a lot more durable than the typical worm gear that only makes contact at one or two points on the thread. More difficult to manufacture so more expensive, but worth it.
I have a thought that chains driving the weight up and down will likely never be an issue. I found an Ansi 40 chain that is rated for working load of 940, corrosion resistant for better life. I called the manufacturer and they claim 4300 lb break point. For 600 lb weight, the working load is 3x above 600lb. There will be a 3/4 shaft connecting the two sprockets driving the weight, so that will not break, I will also tack weld the sprockets to the shaft so I don't rely only on a key and 2 set screws. There is heavier chain I could use that would get over 1600 lb working load per chain.
Upstream towards the motor, if there is a coupler/worm gear/ chain break, that would be bad news. So this gives me the idea to add a break direct to the main 3/4" drive shaft. I have a brushless DC motor with 12.5 : 1 gearbox that when you short the 3 windings together it becomes a very effective break. You cannot turn it by hand with pliars on the shaft. One thought is that when you spin these motors fast enough they will power up and main controller board I typically use, so what about put a bridge rectifier on this motor, and no other controller, and if the weight is free falling faster than the drive motor typically would drive it ( very slowly ) then it will generate enough power to turn on a relay to short the 3 windings together. Sort of a dumb break that could prevent a freefall if the wormgear or drive train failed up stream.
The other idea is I have a shaft break that is 16Nm/2265 oz-in, normally ON = brake and under power the break releases, but this method would require the main Propeller controller to turn it on and off. This would require some other means to determine if the weight is free falling which adds a lot of complexity. I like the idea of the BLDC motor providing it's own power to turn on a relay. The problem is I don't yet know the speeds to know what AC voltage I will get under normal running speed vs freefall speeds. With a bridge rectifier and a large cap, it could force the windings short for enough time to cause a very acceptable freefall to avoid damage. Since it is not above a person, there is less concern but still somebody could lay there head under it if they wanted to so I want to take extra care. I find it quite unlikely to have a breakage up stream including the worm gear.
The photo shows the BLDC on the left, and an electro shaft break on the right.
Overkill if you are using a single chain in the middle, even more so if there is a chain on each side so that should be fine. For a brake you might want to go with something simpler like a spring loaded ratchet that you disengage before starting the up/down drive.
I tested putting a 5v relay on the DC output of bridge rectifier > 5VDC switching reg on the output of 2 wires of the BLDC. At a slow rpm the relay does not turn on but once you start some rpm to represent a anything above running speed the relay turns on and shorts the windings and starts breaking. You may be right about overkill, just needed to consider the options. My main concern is slipping on the keys or set screws. But I think if all the couplers and sprockets are tack welded there is no chance of a failure in the drive chain. That leaves the possibility of failure inside the worm gear, but it is over rated for my application.
I am not finding an electro/solonoid version of this. Have you seen one? I can easily rig up a solenoid to one of these, place it on the output of the worm gear as a backup plan.
I am not finding an electro/solonoid version of this. Have you seen one? I can easily rig up a solenoid to one of these, place it on the output of the worm gear as a backup plan.
I have not seen a solenoid version of a ratchet. I used a standard solenoid and drilled a small hole in the pawl to mount a wire between the solenoid and the pawl.
I wonder if you could mount the solenoid in such a way that the solenoid armature acts as the pawl. Do keep in mind that to move down you will have to move up a bit to release any pressure on the pawl before the solenoid can lift it out of the way.
I have several designs I have made in the past for solenoid latches. Here is one version that mounts on the motor shaft, a solenoid releases a latch. This did have code to first try to release the latch with the solenoid then check to see if a switch was released connected to the pawl. If the pawl was not released, the motor reversed about an 1/8" while the solenoid was still trying to release it. If the switch never gave a true to pawl release the system shuts down.
Upstream towards the motor, if there is a coupler/worm gear/ chain break, that would be bad news. So this gives me the idea to add a break direct to the main 3/4" drive shaft. I have a brushless DC motor with 12.5 : 1 gearbox that when you short the 3 windings together it becomes a very effective break. You cannot turn it by hand with pliars on the shaft. One thought is that when you spin these motors fast enough they will power up and main controller board I typically use, so what about put a bridge rectifier on this motor, and no other controller, and if the weight is free falling faster than the drive motor typically would drive it ( very slowly ) then it will generate enough power to turn on a relay to short the 3 windings together. Sort of a dumb break that could prevent a freefall if the wormgear or drive train failed up stream.
That seems overly complex and would chatter (short then removes relay drive...) ?
Why not flip that, so a relay (no power) defaults to shorted motor, for best low-cost overload speed limiting, and have the power change over when the controller is ready.
If the gearing is such that some creep is possible, than some small motor-hold power may be needed ?
The motor already has the option to brake when PID output is 0. But I don't expect creep anyway through the wormgear. The reason for thinking about the bldc on the main shaft is because the controller will no have a way to detect a free fall under some catastrophic failure. So the bldc braking on the main shaft was to have a system that kicks in during a free fall if something breaks upstrean and speed in the bldc brake is a convenient method to turn on 5v. Yes the relay would turn right back off once it fires but I would prefer a stop start decent over full collapse if that makes sense. I imagine several inches of travel per relay cycle. But it makes things more complex. I will just put a gear and pawl at the output of the wormgear and call it a day. I thing the system is overkill enough with the chains selected.
I will also tack weld the sprockets to the shaft so I don't rely only on a key and 2 set screws.
Some amazing mechanical devices out there; to reduce machining costs, my reducer supplier had me eliminate the keyway altogether and switch to a compression coupling which just slipped on the shaft. I had zero confidence in this thing at first but wow!!! 240,000 LB/IN of torque and an acceleration 1,440 degrees/sec/sec and zero slippage. Yet slacken the compression screws and the reducer would slide right off the shaft.
Well on one hand I like the ball screws too but in this application once installed there is no access to the parts and I have a concern and about rust and maintenance. My philosophy is always zero maintenance. With the chain they could theoretically replace the chain pretty easily down the road if it rusted and some third party could replace the motor if I were not available. I want a 10-20 year life on this as it is a house window. It would be major effort to replace the ball screws down the road.
You could always counter balance like the old school sash windows with the weights. Just use s.s. steel cable with high quality pulleys it would take a lot of load off your drive system.
Without knowing required stroke and other details, I have to say that; for a simple linear translation where gravity is a concern, too many moving parts. Not quite at the Heath-Robinson level but more than enough to have me looking at a small hydraulic power-pack/cylinder/valve. Use a standard bang-bang valve with DC coils and PWM it with a h-bridge, if proportional control is required.
I will do some study on hydraulics today, I have no hands on experience with it. My concern is maintenance and long life with no service. I remember a tractor we had when I was a kid that always had leaks on seals, maybe that is not really a concern these days. The travel is 65". Here is what I have so far with the chain concept. It is very simple. The conditions are that there is no access to the mechanism once in stalled, but there is access to the motor. The motor can only be placed where is shown.
I called a two large hydraulic suppliers just now, both say that you could have a life of 20 years or two days before problems such as leaking. They both said you must plan on replacing the cylinders and plan on how to deal with leaks. I am told there is a concern for under usage as well for drying out the seals when not in use. Basically there will be leaks I am told in my application.
The other concern is space, The total diameter of a device is too bulky. Thanks for the suggestions, always good to explore ideas.
I am typing this, sitting next to a 13 year old machine that I am retrofitting. It runs most of the time but then can sit for a month, waiting for the next production run. Working pressure is 140bar ~2000psi. Not a leak from any of the 12 cylinders and valves. Only oil to be found is the unavoidable slideway lube. I guarantee that you will have more oil dripping from drive chains than you will from a modern hydraulic system.
Danfoss also has "Nessie", which I believe uses pure tap water. I was at a demonstration years ago where the demonstrator filled a drinking glass from the reservoir and proceeded to drink the water.
This is not unusual because with today's quality standards, hydraulic leaks are not tolerated. This factory floor could be mistaken for a hospital. The roller coaster/flight simulators that you find in some malls and amusement parks are hydraulic. Parker cylinders and MOOG servo valves. They are low pressure which is why they are very quiet.
I called a two large hydraulic suppliers just now, both say that you could have a life of 20 years or two days before problems such as leaking. They both said you must plan on replacing the cylinders and plan on how to deal with leaks. I am told there is a concern for under usage as well for drying out the seals when not in use. Basically there will be leaks I am told in my application.
The other concern is space, The total diameter of a device is too bulky. Thanks for the suggestions, always good to explore ideas.
I would likely not oil the chains and use corrosion resist chains. I considered just coating the chain with marine grease and that's it. There will be a chain on each side both driven off the same lower shaft.
Just added a locking mechanism with a solenoid driven pawl.
Have you thought of some safety ratchet or tang on the window slides themselves while it is at rest in any position. A chain or drive failure could result in a huge liability issue in my opinion.
I would likely not oil the chains and use corrosion resist chains. I considered just coating the chain with marine grease and that's it. There will be a chain on each side both driven off the same lower shaft.
Just added a locking mechanism with a solenoid driven pawl.
How would a dry lube work in dusty/dirty conditions, Graphite is hard to keep contained. But it does stick to or embed itself pretty well to materials. Not sure what an application interval would be on your specific project.
There is also chain lubes for motorcycles, bicycles and scooters, that are extremely cohesive, and stick to the chain and sprockets at high speed.
Basically, you need the front end of a fork-truck without the forks. These are hydraulically powered. How many fork trucks are running around factories and warehouses. Imagine the legal problems if staff were slipping on leaked hydraulic fluid. Cars have hydraulic brakes and power steering...how many vehicles are sliding off the road due to leaked fluid. I have forgotten the working pressure of diesel fuel injectors but I know that the lines are tested to 60,000psi.
The point: Leak-free hydraulics are everywhere.
Show me a leaky hydraulic connection and I'm sure I can show you a NPT or BSPT fitting. Dowty washers are fit and forget.
Have a look at a fork truck and maybe borrow a few ideas? Where they do use chain, it's an open-loop arrangement and pulled by the cylinder so stretch isn't a big issue.
So I started out with a ballscrew idea then didn't like that because of potential maintenance and rust/corrosion. Then changed to chain, but similar issues or rust over the years. Then ended up with Gates GT Carbon Poly Chain. 8m Pitch and 36mm wide off the motor to the main shaft, and 21mm wide for the two lifting belts. They claim 800 lb workin load on the 21mm which is 4x the static load. You have to pick a belt partly based on your choice of pulley( minimum radius ) as every belt has a minimum it will bend and maintain it's life.
Here is the progress getting the motor assembly built. Off the motor/gearbox there is 900 oz - in after deducting for 70% efficiency. Then there is a 10:1 worm reducer. Then there is a ~5:1 pulley ratio from a 22 tooth to tooth.. I don't have the efficiency of the worm gear, but the rough estimate so far is 900(from motor) oz in * 10(worm) * 5(belt) 900 * 10 * 5 = 45000 oz in = 2812 lb in. So on the main shaft if I have a 2" diameter pulley it should lift 2812? Anyways I have to lift 500lb.
.... You have to pick a belt partly based on your choice of pulley( minimum radius ) as every belt has a minimum it will bend and maintain it's life.
.. So on the main shaft if I have a 2" diameter pulley it should lift 2812? Anyways I have to lift 500lb.
Interesting photos.
With a need for a large diameter pulley, I wonder if anyone makes a belt+worm drive pulley pair ?
Lifting 500lb is the dead-load, right ?
You need to unstick this (on something rarely used, that could be significant), and then accelerate it, so the momentary load could be much higher.
If it fails to unstick, do you have a sense/cut-out ?
What is the belt failure load ?
Gates will not release breaking load. But each 21mm belt is supporting less than 250 and has a 800lb working load and that is over 3x. The breaking load is much higher but I can't get that data. On other gates-mectrol belts I don't have data for the breaking point is much higher thank working. The motor has current sensing but I have not tested under this much gearing. I could add an idler to the main 36mm belt off the motor with a switch. Not sure what other options I have other than belt tension, the motor can detect overload but at the point it detects it there is already lots of stress!
Comments
Sorry, my bad, you are right, I was wrong. I was thinking of leadscrews, which are dreadful things with sliding friction. Ballscrews have much more efficient rolling contact, as used in the recirculating ball steering box in my '67 Corvair!
Yes, they are a lot more durable than the typical worm gear that only makes contact at one or two points on the thread. More difficult to manufacture so more expensive, but worth it.
http://www.surpluscenter.com/Power-Transmission/Gear-Reducers-Gearboxes/Cast-Iron-Shaft-Input-Gear-Reducers/10-1-RA-GEAR-REDUCER-0-77-HP-RIGHT-OUTPUT-13-133-10-R.axd
This looks to be a good option.
Thinking out loud:
I have a thought that chains driving the weight up and down will likely never be an issue. I found an Ansi 40 chain that is rated for working load of 940, corrosion resistant for better life. I called the manufacturer and they claim 4300 lb break point. For 600 lb weight, the working load is 3x above 600lb. There will be a 3/4 shaft connecting the two sprockets driving the weight, so that will not break, I will also tack weld the sprockets to the shaft so I don't rely only on a key and 2 set screws. There is heavier chain I could use that would get over 1600 lb working load per chain.
https://www.mcmaster.com/#7210k3/=171v906
Upstream towards the motor, if there is a coupler/worm gear/ chain break, that would be bad news. So this gives me the idea to add a break direct to the main 3/4" drive shaft. I have a brushless DC motor with 12.5 : 1 gearbox that when you short the 3 windings together it becomes a very effective break. You cannot turn it by hand with pliars on the shaft. One thought is that when you spin these motors fast enough they will power up and main controller board I typically use, so what about put a bridge rectifier on this motor, and no other controller, and if the weight is free falling faster than the drive motor typically would drive it ( very slowly ) then it will generate enough power to turn on a relay to short the 3 windings together. Sort of a dumb break that could prevent a freefall if the wormgear or drive train failed up stream.
The other idea is I have a shaft break that is 16Nm/2265 oz-in, normally ON = brake and under power the break releases, but this method would require the main Propeller controller to turn it on and off. This would require some other means to determine if the weight is free falling which adds a lot of complexity. I like the idea of the BLDC motor providing it's own power to turn on a relay. The problem is I don't yet know the speeds to know what AC voltage I will get under normal running speed vs freefall speeds. With a bridge rectifier and a large cap, it could force the windings short for enough time to cause a very acceptable freefall to avoid damage. Since it is not above a person, there is less concern but still somebody could lay there head under it if they wanted to so I want to take extra care. I find it quite unlikely to have a breakage up stream including the worm gear.
The photo shows the BLDC on the left, and an electro shaft break on the right.
Thanks for the input guys, most helpful.
I am not finding an electro/solonoid version of this. Have you seen one? I can easily rig up a solenoid to one of these, place it on the output of the worm gear as a backup plan.
I have not seen a solenoid version of a ratchet. I used a standard solenoid and drilled a small hole in the pawl to mount a wire between the solenoid and the pawl.
Why not flip that, so a relay (no power) defaults to shorted motor, for best low-cost overload speed limiting, and have the power change over when the controller is ready.
If the gearing is such that some creep is possible, than some small motor-hold power may be needed ?
Some amazing mechanical devices out there; to reduce machining costs, my reducer supplier had me eliminate the keyway altogether and switch to a compression coupling which just slipped on the shaft. I had zero confidence in this thing at first but wow!!! 240,000 LB/IN of torque and an acceleration 1,440 degrees/sec/sec and zero slippage. Yet slacken the compression screws and the reducer would slide right off the shaft.
The other concern is space, The total diameter of a device is too bulky. Thanks for the suggestions, always good to explore ideas.
Danfoss also has "Nessie", which I believe uses pure tap water. I was at a demonstration years ago where the demonstrator filled a drinking glass from the reservoir and proceeded to drink the water.
This is not unusual because with today's quality standards, hydraulic leaks are not tolerated. This factory floor could be mistaken for a hospital. The roller coaster/flight simulators that you find in some malls and amusement parks are hydraulic. Parker cylinders and MOOG servo valves. They are low pressure which is why they are very quiet.
In such a low-duty usage system, wouldn't you run the chains largely dry - for the same reasons chains are run dry in dusty conditions ?
Does this have a lift-each-side ? - seems another advantage of chains, is they can have a common shaft to keep the lift 'square' ?
Just added a locking mechanism with a solenoid driven pawl.
How would a dry lube work in dusty/dirty conditions, Graphite is hard to keep contained. But it does stick to or embed itself pretty well to materials. Not sure what an application interval would be on your specific project.
There is also chain lubes for motorcycles, bicycles and scooters, that are extremely cohesive, and stick to the chain and sprockets at high speed.
The point: Leak-free hydraulics are everywhere.
Show me a leaky hydraulic connection and I'm sure I can show you a NPT or BSPT fitting. Dowty washers are fit and forget.
http://m.hydraulicspneumatics.com/blog/how-select-hydraulic-connectors-leak-free-reliability
Have a look at a fork truck and maybe borrow a few ideas? Where they do use chain, it's an open-loop arrangement and pulled by the cylinder so stretch isn't a big issue.
Here is the progress getting the motor assembly built. Off the motor/gearbox there is 900 oz - in after deducting for 70% efficiency. Then there is a 10:1 worm reducer. Then there is a ~5:1 pulley ratio from a 22 tooth to tooth.. I don't have the efficiency of the worm gear, but the rough estimate so far is 900(from motor) oz in * 10(worm) * 5(belt) 900 * 10 * 5 = 45000 oz in = 2812 lb in. So on the main shaft if I have a 2" diameter pulley it should lift 2812? Anyways I have to lift 500lb.
With a need for a large diameter pulley, I wonder if anyone makes a belt+worm drive pulley pair ?
Lifting 500lb is the dead-load, right ?
You need to unstick this (on something rarely used, that could be significant), and then accelerate it, so the momentary load could be much higher.
If it fails to unstick, do you have a sense/cut-out ?
What is the belt failure load ?