Notifications
Clear all

Prusa's White Lies  

Page 2 / 2
  RSS
Nikolai
(@nikolai)
Noble Member
RE: Prusa's White Lies

I'm not sure how you guys manage to destroy the original bearings/rods. I keep mine lubricated and clean and so far don't see any reason to replace those.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram

Posted : 23/07/2019 9:28 pm
vintagepc
(@vintagepc)
Member
RE: Prusa's White Lies

Mine were lubricated properly from the get-go. Some bearings must have been crusty, and I had some slightly oversize rods as well. Consider yourself lucky. 

After a binding bearing on Y and then scoring on X, I'd had enough. Also, plastic bushings are a lot quieter than stock metal bearings.. 

Posted : 23/07/2019 9:47 pm
Nikolai
(@nikolai)
Noble Member
RE: Prusa's White Lies
Posted by: vintagepc

Mine were lubricated properly from the get-go. Some bearings must have been crusty, and I had some slightly oversize rods as well. Consider yourself lucky. 

After a binding bearing on Y and then scoring on X, I'd had enough. Also, plastic bushings are a lot quieter than stock metal bearings.. 

True, but wear and tear is also a lot higher with the plastic bushings. Every review so far was saying they need to be replaced more often. Not really attractive to keep the maintenance low.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram

Posted : 23/07/2019 9:51 pm
vintagepc
(@vintagepc)
Member
RE: Prusa's White Lies

Depends what they are. The vesconite ones are relative newcomers and made of fairly hard self-lubricating material, just like the Z-nuts. Time will tell how they hold up, but mine are doing fine.

The stock rods and bearings don't exactly have a stellar track record either...

Posted : 23/07/2019 9:53 pm
Nikolai
(@nikolai)
Noble Member
RE: Prusa's White Lies

I had to replace my z-nuts after around one year. They were worn out pretty badly.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram

Posted : 23/07/2019 10:20 pm
--
 --
(@)
Illustrious Member
Topic starter answered:
RE: Prusa's White Lies

My OEM bearings were reasonable at first. No serious binding or play, mostly they slid easily down the rods though I had to "break them in" and run them up and down several times before the crunchiness stopped.  Crunchiness was probably machining artifacts bound in the preserving oil, but generally good - though I didn't know good bearings until I got the first Misumi's and they didn't have the crunch feel.  Adding grease made the Misumi's even smoother. 

If you print 10 large items you are putting a lot more wear and tear on the printer than 100 small items because small items never reach 200 mm/s.  I suspect this is one reason why there is a significant difference in experience.  I seldom print something under 100 mm. And I maximize print speed choosing infills that can actually use 200 mm/s (line and rectilinear are my primary infills).  I once wore out a set of bearings printing one part, 6 kg of filament in a two week run. lol. Each sub part was close to 200x200 mm, and often over 200 mm tall, with tons of support. 

Posted : 23/07/2019 11:22 pm
Nikolai
(@nikolai)
Noble Member
RE: Prusa's White Lies

Oh yeah. Speed and weight is definitely something will wear out the bearings faster. I would like to see the plastic ones under this stress test. My bet is after 10-20kg filament they would need to be replaced.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram

Posted : 23/07/2019 11:25 pm
--
 --
(@)
Illustrious Member
Topic starter answered:
RE: Prusa's White Lies

For the ultra-curious (and if anyone wants to check my math):

Bearing Life

Posted : 23/07/2019 11:33 pm
Nikolai
(@nikolai)
Noble Member
RE: Prusa's White Lies
Posted by: Tim

For the ultra-curious (and if anyone wants to check my math):

Bearing Life

That looks like making up numbers and not really a scientific approach :). The important data source is missing.

Bearing life in km: all the factors (f variables) are the most important ones and need to be determined. Why are you saying temperature and hardness factor are not important/relevant? Where are all the tables and figures for reference and where did they came from?

Bearing live in hours: ls and n1 are very use case specific. In your case, if you print big things with support, you will have a lot of small/slow movements for the support. Doesn't look to be considered here.

Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram

Posted : 23/07/2019 11:56 pm
--
 --
(@)
Illustrious Member
Topic starter answered:
RE: Prusa's White Lies

Those numbers all come from factual tables derived by the bearing mfg.  So no, not rubbish at all.

https://us.misumi-ec.com/pdf/tech/mech/US2010_fa_p3491_3494.pdf

As for part size and support moves, well, see image above. Those are 50% long stroke moves in the support, and 50mm/s by Prusa default. I set the average speed to 35 mm/s as a conservative matter because I didn't like the life expectancy if I set the value to a real life situation where average speed is 50 mm/s (support at 50 mm/s and infill at 200 mm/s, perimeters at 60 mm/s). 

This post was modified 5 years ago by --
Posted : 24/07/2019 12:14 am
Evan
 Evan
(@evan-2)
Eminent Member
RE: Prusa's White Lies

@tim-m30 How did you arrive at 53 newtons as the applied load? 

For the X-axis, I weighed the stock extruder at around 550 grams (and my BNBSX at closer to 400g), which makes the normal load about 5.39N on Earth. Then that's split between three bearings, so about 1.80N per bearing. There's some lateral load from the extruder's moment arm pulling the top bearing forward (and pushing the bottom one backwards), but not that much.

I never weighed the Y-axis, but I'd guestimate around 25% heavier or so, putting it around 2.25N applied load per bearing, with no twisting effects to generate lateral load. 

Am I misunderstanding what the applied load means in this context? 

This post was modified 5 years ago 2 times by Evan
Posted : 24/07/2019 1:40 am
--
 --
(@)
Illustrious Member
Topic starter answered:
RE: Prusa's White Lies
Posted by: evan38109

@tim-m30 How did you arrive at 53 newtons as the applied load? 

For the X-axis, I weighed the stock extruder at around 550 grams (and my BNBSX at closer to 400g), which makes the normal load about 5.39N on Earth. Then that's split between three bearings, so about 1.80N per bearing. There's some lateral load from the extruder's moment arm pulling the top bearing forward (and pushing the bottom one backwards), but not that much.

I never weighed the Y-axis, but I'd guestimate around 25% heavier or so, putting it around 2.25N applied load per bearing, with no twisting effects to generate lateral load. 

Am I misunderstanding what the applied load means in this context? 

I was focusing on Y since those are the bearings that fail first on my printer.  The calculation uses dynamic loads ... and I guesstimated based on what 10 pounds will do. 

Load Factor (fw)
To calculate load applied to the linear system, in addition to object weight, it requires inertia force attributed to motion velocity or moment loads. Further, it is necessary to accurately determine the temporal change of each. However, for reciprocating motion applications, it is difficult to obtain accurate calculations due to the effects of vibrations and shocks. Therefore, use the Table below in order to simplify the Life Calculations.

And don't confuse a static load spec with a dynamic load spec. They are significantly different, Misumi's are about 380 static, too.

Also, reading the paper, loads being distributed doesn't really affect the life much - distributed on two bearing it isn't 50%, but a 19% decrease. Three drops it a bit more to 28% reduction in wear. But the Prusa design floats half the Y weight on one bearing, so I went with the 81% figure to average things. 

So whatever the full Y platter weighs, add 750 grams of plastic, and you probably have a good static weight to work with. Then add some dynamics like power cable torquing the bed, the 2:1 bearing losses adding more rotational torque, and ... well:  I went simple and guessed at 53 N, 20% of the dynamic load spec - about 11 pounds.   It might be half that. 

Even if the dynamic load is off, printing at high speeds like I do changes my printer life of about 250 hours instead of the 1470 shown, even quartering the dynamic load isn't that helpful. 

 

But then running through the math again found a missed cubed factor 😫 ... if I believe the result now, the life of the bearings is 11.3 years - instead of 370 km, the base is now 35.6 thousand km...  or 100,000 hours.  It also makes me wonder - if folks can expect over a million km - why, in their example do they only use a 50 km life.  lol.  Interesting mental exercise since experience and bed wobble dictate when I need to swap bearings - and so far has been about 3 to 6 months (and would be more frequent if I didn't use my rubber band to hold it together longer).  

 

This post was modified 5 years ago 3 times by --
Posted : 24/07/2019 3:23 am
--
 --
(@)
Illustrious Member
Topic starter answered:
RE: Prusa's White Lies

New bearings and rods: $192 ...

Seems I messed up last round when I tried LMK instead of LMU bearings. I didn't see that despite having two more rows of balls the LMK is rated for half the load of LMU.  Hence, less life than expected. Shooting one self in the foot for having a "better idea" can be painful.

Ordering high precision hardened chrome rods this time around, and metal carriers in the LMU bearings rather than plastic just in case the 80c limit of the plastic carrier is an issue on Y.  

Posted : 25/07/2019 10:48 pm
Page 2 / 2
Share: