PhD topics?

I've decided to do a PhD and for that you need to do 'original research'. The only thing I can think of is 'something to do with Reprap'. I've asked a few people who are interested but not involved with the project and they've given me a few directions to go in but nothing that really pushes my buttons, so I'm asking you, the people intimately involved, what is there that is worth spending the next 3 years of my life researching that would be helpful to the project?

I think what I need is something very narrow and specific that the commercial world wouldn't want to spend the time doing.

I'm primarily a software guy (more gluing things together rather than a programmer though) and my passion lies in making Reprap more user-friendly and accessible to the general public. Although if you have any other ideas, feel free to discuss them as well. Maybe someone else reading this will find its something they want to do.

Maybe out of this will come a short list of prospective PhD topics that people can cherry pick from and we can have a page of current research efforts on the wiki.

I have some experience in managing PhD level research. I think Adrian has more than I do. For myself, I'd be hard-pressed to suggest a PhD topic till I got a look at a detailed resume of your background and capabilities. I could suggest all kinds of things, but it would be shooting in the dark without knowing what kinds of analytical skills and coursework background that you have.

Basically, I need to know more than "I'm primarily a software guy (more gluing things together rather than a programmer though)".

The general rule of "original research" means that you more or less have to do something from scratch. Mind, you may use extant tools to do it but the concept, at least, has to be new.

Just off the top of my head, building a good numerical model of thermally-driven warping in prints would be a hell of a good topic. The literature I've read on the subject has been very hit and miss. I've thought about doing it, but I don't know of any software apps that let you model the thermal behaviour of an object as it is constructed. The ones that I've used like Fluent, for example, generally assume that the object already exists in toto at the beginning of the simulation exercises.

Hi Reece,

... some intersting topics (as for me) would be:
- 'room-temp-fabbing' with/without curing the slices
- serious paste-'alchemy' or developing 'green' fabbing materials
- noncartesian (e.g. parallelkinematic) robots
- 'self-optimization' (e.g. with neural-network-simulation)
- fabbing with prefabbed blocks ...

Some of this themes are already made elsewere, but there is still so much work to do

Viktor

Wow, when I read your post my mind immediately jumped to something along the lines of what Forrest suggested. I second that. It would be really useful - if we can consistently model warping behavior with software (either completely virtually or through an evolutionary but still computer controlled trial-and-error approach), it should be possible to compensate for it and produce even more accurate models.

Well because like everyone else , we dont have much to go on i will throw in an idea. You could work on an plotting efficacy. What I mean is that offten the print head back tracks or takes the long way. Such as printing the "widebar.stl" after if finishes the first layer infill it could move or a mm or 2 and start the new layer boarder, but it instead goes back to the staring position from the first layer which is some 50 mm away. Besides lost time there is also the fact that the nozzle is still oozing all that way. So yea not the easiest thing to do. Considering it would come down to having to analyze any and every stl think of the fastest way to start the of a new layer. Of course this is all kind of assuming that this is in your field.....

I would also ask on rp-ml (rapid prototyping mailing list):
(Subscribe by emailing "[email protected]")
The server is having a hiccup, but normally is up:
[rapid.lpt.fi]

I would also take a look at "Rapid Prototyping Journal"
[info.emeraldinsight.com]

Along with
[www.tctmagazine.com]
(The magazine seems to be free if you give them personal details.)

Possibly
[www.timecompress.com]

and
[utwired.engr.utexas.edu]
who host the annual
Solid Freeform Fabrication (SFF) Symposium

By the way, what department would you be in, and what kinds of mentors and resources would you have access to?

A suggestion: Computational Geometry thesis - whip up an algorithm and user-friendly software to do object reconstruction from a bunch of digicam photos. (Such software exists, e.g.
[ezra.dev.java.net]
but the world could use more.)

Sorry I haven't replied to these responses very fast, I've been quite busy at a paying job, next year will be a different story :-)

I was thinking something along the lines of adding 3D scanning ability but it may be that Sebastien's idea about object reconstruction from digital photos would be easier. I'll have to look into it.

I realize I didn't really give you much to go on as far as my specialties go but I was also just wanting to get a list of all the problems that need solved that might take a bit of R&D and I think thats what I got.

Thanks for the replies. I'll talk it over with my potential supervisors and we'll see what happens.

reece.arnott Wrote:
-------------------------------------------------------
>
> I realize I didn't really give you much to go on
> as far as my specialties go ..
>

Yeah, you sure didn't.

Edited 1 time(s). Last edit at 08/24/2008 02:32AM by Forrest Higgs.

Reece,

An option derived from Sebastien's suggestion to whip up software to do object reconstruction from a bunch of digicam photos, would be to make software to turn pictures of someone's feet into an orthotic support model, which the reprap would then fabricate.


Ronanwarrior,

I don't know what software you're using to print, but skeinforge does try to find the shortest path; and if it doesn't find it, that's a bug

Well okay it does try but at the same time it limits it self. Like when making a circle it always has the same starting point. It does switch going CW or CCW but it never starts at a new location. Besides taking time to get to the seem starting point it can also lead to other problems. It can make it so flaws are consistant through out the part. In my case it leaves a little bit of a blob at the start of each circle. This later become a long ridge in the hole. If it had stagger out the circle boarder it would mean the "error" is one less pronounced and now easier to get rid of by lighly running a drill through the hole. It not a horrible problem.

Thanks Ronanwarrior for describing the problem. I'll add loop entry staggering code within a couple of months.

Cheers,
E

ronanwarrior Wrote:
-------------------------------------------------------
> It can make it so flaws are consistant through out
> the part. In my case it leaves a little bit of a
> blob at the start of each circle. This later
> become a long ridge in the hole. If it had stagger
> out the circle boarder it would mean the "error"
> is one less pronounced and now easier to get rid
> of by lighly running a drill through the hole.


Good point. I noticed something similar on the minimug I printed. There was a seam running up the side.

Yes I think actually closed loops should not actually close but end a bit early otherwise there is an excess of material which causes the seam.

---

[www.hydraraptor.blogspot.com]

With some of my laser cut parts, I specify exactly where I want the cut to start/stop. There is always a blip there and there are parts where it would be a problem if it occured randomly.

In FFF it would also be nice to be able to specify where the seam is. Random by layer may work for some parts, but would make a pimply mess of a beautiful flat part. I dont think it will ever be fully solved, but having the algorithm place the join in randomly selected concave sections may help.

Or, it could look for the tightest concave angle in each profile.

Holes could be designed with a small notch running up them, which the software would see as the tightest concave segmets and automatically use to start/stop in - which would reduce the need to ream out after. The designer could specify this as a spiral notch.

The only problem may be that the start/stop would end up on the highest tensile load point of the part with obvious structural impact. A user friendly opton in the host software would be - slice for strength or slice for asthetics.

This is where a milling pass after every FFF layer would solve all the problems. The milling bit would only have to be as long as one layer is deep, so could be very fine giving good feature sizes and very high accuracy. It would fix all the extruder overrun issues as well.

---

[www.hydraraptor.blogspot.com]

I've decided to do PhD on HR as im working as a lecturer.please help me by giving some new topics for research.

thanks

i have decided to do phD in HR. so pls help me out by suggesting some latest topics as soon as possible

bhutani, I'd suggest you start by building a machine. And also, I'd suggest you spend a few sentences introducing yourself, and mentioning what you find exciting about RepRap, what you plan to make with your machine, and how you are building a machine.

Otherwise, no one is going to spend time helping you with this matter, since they'll get the feeling you are trying to use peoples' time as a service rather than speaking to fellow people in a community.

Hi,

Can you please suggest me some excellent HR topics to do phd in software company.

Regards,
Jyothi

Hiii,


M working as a lecturer in management dept., nd interested to do P.hd. in Hr, can u pls suggest me topics 4 dat.

regards

What about this one:

- "Evolving from the motivation of an ideal to reality and then creating progress: dynamics of creative individuals"

Or this one:

- "Why motivated people stop working on putting ideals into practical solutions"

This one would obviously try to understand why certain people do NOT follow up on until the end after an initial implication in e.g. Reprap building.

hai am amar anth reedy am qualify in r cet exam so pls suggest me for suitable phd topic for me and also i have 2 years industry experience as a hr officer.

amar, I'd suggest you start by building a machine. And also, I'd suggest you spend a few sentences introducing yourself, and mentioning what you find exciting about RepRap, what you plan to make with your machine, and how you are building a machine.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

Peer, what about "The Fatal flaw of Consensus: Debordist image and the Subsemanticist Paradigm of 3D printing" (via: [www.elsewhere.org]) , or "Derrida and a Stakeholder Driven-Search for Excellence in Self-reproducing Systems" (my own humble creation)?

I believe either of the two subjects might be excellent human resources thesis topics.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

A gentle reminder:

Students, please note that RepRap is a 3D printer project.

As such it has a lot to do with art, engineering, design, humanitarian causes, and possibly chocolate (according to wikipedia).

Mansi, Parika, Nishtha, stuff like this http://reprap.org/wiki/WolfStrap-English has little to nothing to do with human resources.

---

Engineering, Design, and Art students, please feel free to start new threads introducing yourselves. This thread is a little strange. It's the 20th hit on google for "PhD topics on HR", which is why HR PhD students started coming in here.

I very rarely lock threads and thereby censor speech because that's not how we do things here at RepRap, but I'm going to lock this thread. This way Peer and I won't get too mischievous, resulting in some poor student spending six months trying to do a thesis about Mendel as a HR subject.

Edited 1 time(s). Last edit at 07/24/2010 02:06AM by SebastienBailard.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!