Maybe a dumb question

Hi there all

I have a Rapman 3.1 and starting to build a Pursa Mendel. My question is, how do u make the X axis motor mount without useing any support material. (if you use rapman u have 2 materials to make this)?

Some part are hanging in the air.

Maybe a dumb question but i have ti ask.
//Janne

This is not a dumb question. I think the part is supposed to be printed with your bridging parameters adjusted correctly, so the printer speeds up across the gap eliminating the sag. It helps to have a fan also cooling the part during printing. I looked at this part and saw the same issue as I haven't had much success with bridging setup in skeinforge.

This is why I redesigned all the parts in my prusa build to simplify printing.

Ahh, tnx. Looks like a new cad job for me then. The sad part with RapMan is the time its takes for support material to warm up for every layer u make.

Yes, I had the same reaction. This part does print with a reasonably well tuned machine, but bridging especially at that scale has been reported to produce weak parts, and is a rather useless kludge. Bridging should be avoided to increase printability for all. Have a look at Adrian Bowyers build notes in the wiki, there are a couple of redesigned parts there, without the bridges as far as I remember. Not sure if they are for LM8UU, though.

martinprice2004 Wrote:
-------------------------------------------------------
> I think the part is supposed to be printed with your bridging
> parameters adjusted correctly, so the printer speeds up across the gap eliminating the sag.

To adjust the bridging parameters, you need to reduce Speed/Bridge Feed Rate until the ends of the bridge filaments adhere properly, and then reduce Speed/Bridge Flow Rate until the bridge filaments stop sagging, and pull up tight.

It takes some effort to adjust, but it makes things possible that otherwise would not be.

Lanthan wrote:
-------------------------------------------------------
> bridging especially at that scale has been reported to produce weak parts, and is a rather useless kludge.

I have to disagree. The x-motor ends that I have produced are quite strong, even at the bridged areas. "useless kludge" is quite an over-reaction.

One thing I have found is that slic3r handles bridges better than skeinforge/sfact. skeinforge applies bridge extrusion settings to every last bit of a layer, if it detects a bridge anywhere in the layer - even if the layer is made up of a large number of separate islands. slic3r applies bridge extrusion settings only to the area that is being bridged.

raldrich Wrote:
-------------------------------------------------------
> martinprice2004 Wrote:
> --------------------------------------------------
> -----
> > I think the part is supposed to be printed with
> your bridging
> > parameters adjusted correctly, so the printer
> speeds up across the gap eliminating the sag.
>
> To adjust the bridging parameters, you need to
> reduce Speed/Bridge Feed Rate until the ends of
> the bridge filaments adhere properly, and then
> reduce Speed/Bridge Flow Rate until the bridge
> filaments stop sagging, and pull up tight.
>
> It takes some effort to adjust, but it makes
> things possible that otherwise would not be.
>
> Lanthan wrote:
> --------------------------------------------------
> -----
> > bridging especially at that scale has been
> reported to produce weak parts, and is a rather
> useless kludge.
>
> I have to disagree. The x-motor ends that I have
> produced are quite strong, even at the bridged
> areas. "useless kludge" is quite an
> over-reaction.

Useless kludge is the word, I am not over-reacting. That particular piece is a good example, the bridge is the lesser of the problems with it. I like to tension the belts as recommended for good accuracy. Well (and I am certainly not the only in this case), the belt tension would flex the X motor mount wall so the pulley axis would not be perpendicular to the X axis anymore, and the belt would slip up the pulley. I still have in place the emergency bracing I had to resort to. Other people resort to reinforcing the motor mount wall with aluminium profiles.
I have a redesign of the X axis in the pipeline. It will be really ugly, boxy, in three parts, and it wont flex at all.

Still considering the X ends: the designer maybe forgot that he was dealing with very thin layers of deposed filament when placing the nut traps under the X ends, so no way to really secure tightly the X smooth rods. That one was solved by drilling and taping M8 nuts as rod collars.
I could go and on (who had the oh so brilliant idea of hanging the whole X axis from the Z motor couplers? Do I have to resort to the barber shop and hair spray as Triffid Hunter recommends to have those couplers stop slipping from the aquarium tubing and the motor shafts?)

I can see (and undestand) many reasons why some people would want to take a machine design and radically slim it down. But as a maker, I feel that my time and peace of mind are more important than a few more grams of material.

So why did I not print the Selis mendel instead? Not all in the simplified design is bad. And the information on the "simplified mendel" design shortcomings only appeared gradually here in the forums. Yes, the pieces certainly printed faster, but I spent a good more time dealing with the... yes, kludges. This stuff must still evolve, and it will. I am more careful when considering a new design now.

>
> One thing I have found is that slic3r handles
> bridges better than skeinforge/sfact. skeinforge
> applies bridge extrusion settings to every last
> bit of a layer, if it detects a bridge anywhere in
> the layer - even if the layer is made up of a
> large number of separate islands. slic3r applies
> bridge extrusion settings only to the area that is
> being bridged.

If one forgets what happens with the rest of the layer, in terms of quality, skeinforge's bridges aren't bad at all.
slic3r isn't quite mature enough, right this evening I had it transforming circles in spirals going out of the (very simple) piece. Will post a pic in the appropriate forum. The issue is probably with the circle detection code.

@Lanthan:

In the context of the original post, your stance on bridges is an overreaction, although I agree with you that not all of PrusaJr's changes to the design were well advised.

My concern about the Prusa parts repository is that more and more it requires a finely tuned printer and highly knowledgable operator to print usable parts.

In particular, the "Friction fit" x-ends can be tremendously difficult to print.

As for your slic3r issues - curve detection and conversion is very new, and still not even beta quality - errors there are to be expected.

raldrich Wrote:
-------------------------------------------------------
> @Lanthan:
>
> In the context of the original post, your stance
> on bridges is an overreaction, although I agree
> with you that not all of PrusaJr's changes to the
> design were well advised.

My point was that: to warn a gentle newcomer that bridging may not be the only (nor main) issue with this design.

>
> My concern about the Prusa parts repository is
> that more and more it requires a finely tuned
> printer and highly knowledgable operator to print
> usable parts.

I fully share this concern.

Maybe the Prusa Mendel is ideally targeted to a public of knowledgeable, experienced parts & kit resellers.

However, a quick look at ongoing work in new designs (Nophead, Kludgineer, Buback... as well as most of the work on the vertical X axis, etc.) shows significant improvements on all of the mentioned points. Knowledgeable and advanced users are routing around problems and biases, and this is very positive for all. There appears to be wide awareness of the issues with bridges, and I am not seeing any overuse of them in the new designs

> In particular, the "Friction fit" x-ends can be
> tremendously difficult to print.

and they require precse dimensions in smooth rods. There was some quite nice research behind it, but the old clamp solution is still superior.

> As for your slic3r issues - curve detection and
> conversion is very new, and still not even beta
> quality - errors there are to be expected.

found a way around it - of sorts. I am not complaining, I am in awe looking at the evolution.

One shouldn't judge other peoples designs so harshly. We can all see improvements in a design its called "20/20 hindsight"

I would say however that if you can avoid bridging and support material and avoid overhangs of more than about 60 degrees it will be printable by anyone even if they are struggling with setup which can only be a good thing.

All the parts on a prusa and wades extruder can be printed without support and bridging with a little redesign.

My iteration of the prusa proves that here and I am sure there are many others that do the same.

martinprice2004 Wrote:
-------------------------------------------------------
> One shouldn't judge other peoples designs so
> harshly. We can all see improvements in a design
> its called "20/20 hindsight"

Yes we all make errors but... it took nearly a year for this stuff to fully surface, and many people keep copying and reproducing parts with the initial flaws. One could attribute that to "imperfect information". Moreover, many of us are reprapping with long interruptions.
The question behind this is: can we escape the infinite monkey theorem and streamline the widespread lesson-learning that is taking place?
People in the scientific community have "peer review" (yes it is not that obvious in our times of hyper-specialization, but most of the time it does its job), could we imagine an equivalent for repraps & repstraps?
Peer review would not be so easy for printable machines, since to test a design the would-be reviewer has to invest time, energy and plastic: justifiable when one needs a new machine, and then, "supposedly proven" designs get first attention. But what about encouraging designers to provide more information and discuss their choices? (yeah they will find it tedious... designing is fun, as for documenting...)

>
> I would say however that if you can avoid bridging
> and support material and avoid overhangs of more
> than about 60 degrees it will be printable by
> anyone even if they are struggling with setup
> which can only be a good thing.

A page "recommendations for designers" in the wiki?

>
> All the parts on a prusa and wades extruder can be
> printed without support and bridging with a little
> redesign.
>
> My iteration of the prusa proves that here and I
> am sure there are many others that do the same.

You sure put a lot of work there.
How to reinject design improvements in the mainstream?
Then we find the many obstacles: file formats (parametric openscad files vs. propietary 3D design packages, availability of sources...)

It does take time, money, and plastic to print out someone elses design, and it takes even more time to edit the wiki, write a review or take pictures.

Often, the people experimenting with the new designs are those just getting into the hobby, and so have the least experience (in terms of reprap printing. they might have oodles of relevant experience, though, like engineering degrees, etc).

the best way to encourage people is, unfortunately, money, which we don't have much of. But if a new designer wants to offer rewards to people who contribute significantly, that might get the ball rolling. A good example of this working is the vertical x contest.

Another thing we can try is "achievement badges" like they are doing with video games these days. When i was in boy scouts, we used to get patches for going to certain camporees. they've use similar patches for spaceflight missions, and militaries have their badges for various campaigns. I don't know how this would be organized or administered, but if it is interesting to anybody, i'll make a new topic in another forum and we can work out the details.

I think a badge system could be great idea. I would guess an achievement system like slashdot has would be pretty effective. And I think it would need to be automated as in slashdot so that it doesn't just die from admins neglecting to update it.