Innovative Production Machines and Systems Conference

There several ways to extend tungsten tool in traditional cutting process:
(1)Proper geometry of tool;
(2)Considered process parameters;
(3)Suited material which include appropriate grain size of tungsten and compatible Co contents, it can ensure the suited hardness and wear resistance, as well as the toughness and impact resistance.
For this specific Micro-EDM process in your case, I am sure the tool material should be a kind of fine tungsten carbide, which has the tungsten grain size less than 1 micron, because it can one print head (260 holes) from one electrode.
There are two ways could be considered to extend the tool life, the first way is to optimize the parameters of this Micro-EDM precess, such electric current and impulse intensity, as well as dielectric type so on and so forth; the second way is the improvement of tool material, more finer grain size of tungsten should be an effective way to extend the tool life, so far, the tungsten grain size have reach to 50nm, it may provide high preformance for tungsten tool, but the Co contents is another important factor need to be considered.

Hello Paul and Prof Xiuxu,

I'd like to contribute with a few comments and questions.

Questions:

A)
What are the quality functional characteristics of the holes that you are after? That is: roughness, error of form, error of dimension, error of position, mechanical strength/small or absent heat affected zone etc.?

B)
Which of the following are you interested the most:

1.reducing to one half the cycle time (time for producing a part).

2. Increasing the productivity by reducing the set-up time.

3. Increasing the efficiency of the use of the electrodes by doubling the productivity of one electrode.

C)
Do you drill one hole at a time or you have a shaped tool which drills all the 260 holes in one go ?

Comments:

For micro-EDM operations, I am aware of commercially available electrodes either in W or in WC. In this second case the grains of WC are bounded together by Co.
Certainly, it is expected that considering different tool electrode material (from W to WC for instance) may significantly increase the tool life (see the others paper in this section for an idea of the effect of the material on the wear ratio). However, these electrodes are down to 90 micro-metres diametre. They may have introduced smaller diameters nowadays, I don't know. If you want them to be smaller you need to EDM grind them.

Also, the flushing plays a germane role. So, again there are commercially available pipe electrodes (i.e. with hole through the electrode) for flushing or sucking through the hole. But this, down to 120 micrometer diameter, again, if new models have not been introduced meanwhile. Or again, you might considering rotating the electrode at relatively 'high' rpm while machining. Do exists ad hoc spindles to attach to the machine for that.

If the intent is to halve the cycle time and you are drilling one hole at a time, why not considering to design an holder and or tool for producing at least to holes at a time? Or did I get the main problem wrong?

Best regards,

Carlo

"Multi famam, conscientiam, pauci verentur." (Many fear their reputation, few their conscience. -Pliny, Letters)

Hi Carlo and Prof Xiuxu,

Thanks for the replies,

Unfortunately I dont really know much about the first set of questions that you asked about Carlo, ie roughness and mechanical form, but the holes have to be drilled so that there is a consistent pitch between them (measured to within a micron) and they are measured for roundness at the entry point of the electrode by a microscope and computer software. They are visually inspected for roundness at the exit point of the electrode.

In regards to the other questions,

We drill one hole at a time.

and

I am mainly interested in Increasing the efficiency of the use of the electrodes by doubling the productivity of one electrode.

As far as WC goes, I have looked into this and other materials but unfortunately it is not available at such small diameters.

I will look into the pipes but again I feel that the diameter that I seek will be too small.

We are using rotating electrodes with the spindles already but thanks for the suggestion.

Making a tool that has two or more spindles could be an idea but I dont think I would be given the resources to do this to be honest..

Do you know of anybody using EC-CVD diamond at diameters like 40um? I have come across a limited no. of papers on the subject but none that will give me a definative solution for what I am trying to acheive. The results achieved in the papers did sound promising and I was wondering if you have heard about anybody using this technique?

Thanks,
Paul

Hello Paul and Prof. Xiuxu,

Thanks for the answers. From them, I argue that positioning error (the pitch) and form error (roundness at the entry and exit points) are the sole functional quality characteristics of the part. If others had been prescribed by the product designer(s), they would certainly be undergo a measuring process. My question was meant to have an idea of the objective function that might be potentially used in a parameter optimisation as suggested by Prof Xiuxu and to have a better idea of the contraints that may drive a potential consideration of manufacturing processes alternative to micro-EDM.

If you are already rotating the electrodes, it may be even more difficult to implement tools that drill more than one hole at a time.

Unfortunately, you are the first person I hear about ED-CVD electrodes from. I've made a quick search and quickly read http://www.sodick.co.jp/tech/img/article_s06.pdf. It looks quite promissing and iteresting.

Another comment: What about the dielectric? Have you considered investigating the effects of different types of dielectrics (different types of hydrocarbons and or deionised water)?

All this discussion generated a few specific ideas in my mind. If you wish, they may be the source of potential collaboration involving also Atanas. He can significantly contribute to make this ideas real or disprove them (he knows). He is also in a more stable position. I am going to contact you via email.

Meanwhile, I was also wondering whether Prof Xiuxu can perhaps give us his impression on how micro-milling could potentially constitute an alternative to the Paul's current -in-use micro-EDM process.
Do cutters for directly milling such holes exist? Or could they be somehow produced? And if they were, how many revolutions per minute should the spindle rotate to provide an appropriate cutting velocity?
Alternatively, could micro-milling be used for cost-efectively producing micro-EDM electrodes for this particular operation? Any advice of how to handle and move the machined micro-EDM tool from the milling machine to the micro-EDM machine with limited risk of breakage?

I look forward to reading both of your answers and/or of whoever wishes to post their most welcome comments.

Best regards,

Carlo

"Multi famam, conscientiam, pauci verentur." (Many fear their reputation, few their conscience. -Pliny, Letters)

Hello Carlo,
I have to apologize for my late reply, because we hold an exam in the past two days. And I failed to upload my video file for its large size.
I am very pleased to read your comments about the problems in micro-EDM drilling, I am sure you know a lot of tips about micro-EDM. When you talked about micro milling, I have a friend doing the research in micro-milling, and we have talked about micro drill when we worked together. As we know, the drill tend to be more brittle in the maching process, especially in the interrupting process, such as milling. Because the cutting forceand the temperature in the cutting area will be no-consistent, and it will be the factors to lead the tool failure. I think it is also the reason they use micro-EDM to drill the holes in such a small size. So, I agree your suggestions.

Dear Carlo,
Thanks for your so many good questions!
The first question: actually, we have done this material analysis on five fractured tools, at two specific locations：tip and shank, and the results have shown the same tendency. We did think about the impact of the WEDM on these specimens, so the process of samples preparation need to be cosidered in detail, either the parameter of WEDM or the polishing process. The peaks of the X-ray diffraction results of Figure 3 and 4 show the specific elements in this specific location of tool material. As a non-destructive method, the X-ray diffraction usually are used to determine the constitute elements of material qualitatively.

The second question: we have done a serial test to clear the variation of Co contents in the tip section, we use the EDS line scan in SEM, it a kind of method to detect the constitutes point by ponit, and the step can be set up in SEM, such as 0.1 micron. The results shown there less Co in the surface layer and there are more Co in the center of the section. It was caused in the titanium machining process, because titanium has high chemical affinity.

The third question: the use of statistical data analysis in performing experimental work in manufacturing research at University of Michigan are wide-spread, as I know, they have done a lot of project for comerial companies such as Big Three( Ford, General Motor, and Kreisler)and other well-known companies. Unfortunately, the use of statistical data analysis in performing experimental work in manufacturing research at Wuhan University are not wide-spread. Right now, a SPC project is conducting for a manufacturing company in my group, it is mainly to control the quality and improve the manufacturing process capacity. We may discuss a little bit on it if you have interesting.

Thanks again for your kindly attention!
Best Regards,
Xiuxu

Dear Prof. Xiuxu,

I am grateful for your detailed clarification about my previous post.

My first question stemmed from the fact that 5 tools have been explicitily considered in section 3.1, whereas it was not clear wheather Fig. 2, 3 and 4 where one of the 5 test results actually performed or where the sole test performed. Your excellent clarification dissolved any doubt.

Question two: if I have well understood, the affinity of Ti with Co makes the Co 'to stick' on the Ti chips in such a way that the surface of the tool becomes more brittle than the core due to a decreased Co content. This is the suggested assignable cause of the tool failure. Would you therefore suggest any other binding material for the WC that has less chemical affinity with the Ti than Co has? Are there other potential cutter material already available that may be worth comparing and contrasting with Co-WC for machining Ti alloys?

I am glad to hear that at least somewhere there is some both academic and industrial use of quantitative methodologies. Sometimes, I had the feeling that nobody cares about it at all and that I have consequently wasted a long time for nothing (the same applies for micro manufacturing). Should you like to tell me more about the SPC project that you mention, I would certainly be glad to discuss it further.

Thank you for all of your detailed posts at the IPROMS2007-micro and nano manufacturing session.

Best regards,

Carlo

"Multi famam, conscientiam, pauci verentur." (Many fear their reputation, few their conscience. -Pliny, Letters)

Dear Carlo,
I have to apologize again for my late reply, please forgive me.
I appreciate your question about whether there is any other binding material for the WC that has less chemical affinity with the Ti than Co, it's a very good question. There are already some works have been done at this point, and the conclusion is Co is the best binding element for WC. In 5th international high performance cermic, which was hold in Changsha in May 11 2007, there is a French scholar show his research on finding other alternative element for WC, but so far, the results show he only can find some element can act the same as Co. For Ti, another problem is most of the coatings are not as effective as other workpiece material, such as Ni.

Thank you so much for your interesting in the SPC project! So far, we have develop a software platform for our sponsor, which can collect the data in the production line, and do the realtime analysis, and calculate the process capability. Because the data is measured manually in this system, so we are considering to use the digital measure equipment to get the data automatically, can you give us some recommendation about this kind of digital measure equipment, the data we measured are mainly the leangth and the diameter and position error. I am alos looking forward to get your valuable suggestion the control method of the machining process.

Thank you for all of your detailed posts on my paper.

Best Regards,

Sincerely yours, Xiuxu

Dear Prof. Xiuxu,

Thank you for your answer.

Due to the fact this virtual conference has been officially closed and I don't check this website as regularly as I used to and also due to the fact that I don't know exactly until when they are going to keep this site working, I would suggest to continue our discussion directly via my email contact address. When they will close also that, I will let you know another one. But at least, I know when they are going to close that address.

I understand that finding a binding component for WC tools is quite an active area. I know too little to add something more.

About the SPC, I have a few questions and some comments, as usual I lack of answers, but what can I do? I know very little.

1. I don't think I have enough in-deep knowlege of what equipement is available out there also I don't know sufficient details of your project for saying something specific. Anyway, as a very general personal attidute I see two different main stream:

a) (the one I dislike) To rely on some company as supplier of integrated measurement systems: but in this way, the company usually hide all the algorithms, all the detail, all the explanations. Your just buy from them, they get your money and forget about research and trying to understand and test different competing solutions. I am sure you know many of these companies, don't make me advertise them for free.

b) Do it in house on your own, so you can endeavour to know, although it costs a lot in terms of efforts, time and struggling. Upon this, I would say that I am quite favourable biased towards GNU/Linux and Free software in general. So I would give a look at http://www.comedi.org/ and http://rtlab.org/index.php and http://rtxi.org/index.jsp and https://www.rtai.org/ . The last one is perhaps a bit too far, but it may have some interesting application links. It is just incidentally that is from my old school in Italy: I discovered here in UK and I didn't have any part in it, i.e. I know nothing about all this.Beside, I've never used any of these up to the completion of something. Take it as a 'disclaimer' to protect myself if somethig went wrong in your tries ( :^) )

c) I would not underestimate the necessity of having available traceable reference materials/ artefacts for applying the ISO 11+something about linear calibration and control of measuring systems.

2. As a general attidute I would start considering as input parameters of the project, what I believe both the company and the researchers in the area value the most, namely, the costs. How much are the cost for searching for assignable causes, producing in out-of-control-state, and sampling from the process, just to name a few? Is there a cost model suitable for such considerations? Is the choice of an automatic measurement system rather than a manual one justified on the basis of all the cost involved?

3. From your words I argue that the quality characteristic of interests have already been identified via an ABC/Pareto analysis or some something else.

4. Also I argue that you are exclusively interested in Capability studies, that is to assess if the variability of your process is suitable to the specification limit of your product. Have you already identified the probability distribution function of the quality characteristic? How? And what exactly would you like to monitor the Range, the standard deviation, the variance, the PCR, others?

5. "the control method of the machining process" This sentence is not entirely clear to me. I know, I am quite limited. Are we talking about the design of control charts for this (these) quality characteristics (length, diameter, position) or a problem of control of the tool motion in (micro)-EDM ? Or both :^) ?

6. Also, "real time analysis" require a bit of specification for me to understand. Do you mean On-line quality control or are you actually talking of 'soft real time' or 'hard real time' and similar topics?

7. About the software for statistical analysis, Earlier on I was using Minitab whereas, again recently, I have been training myself to using a free software tool, R ,from www.r-project.org (both for linux-like and windows). It looks pretty good, although a bit complicated. But I believe it has and it is forcing me to learn more. In there there is also a package (qcc) for SPC based on Montgomery. And again, by coincidence come from an Italian guy. Please, don't misunderstand me, I don't know him, and I am not trying to advertise fellow countrymen. They are really just curious coincidences for me. That is the sole reason why I am mentioning them.

Sorry, that I've been a bit long. I hope it might help.

I look forward to hearing from you (email perhaps better, as I said).

Best regards,

Carlo

"Multi famam, conscientiam, pauci verentur." (Many fear their reputation, few their conscience. -Pliny, Letters)

Hello,
I can see no fundamental reason why one cannot double the number of holes achievable with one electrode.
However, as we have not done this yet, I cannot guarantee success.
I am sure our colleagues at the MEC will be more than pleased to investigate this possibility if you would like them to do so.
Please feel free to email us, initially, at manufacturing at cf.ac.uk

Hello Mohd,

I have seen your message just now. Had you made your email address accessible in your profile, I would have contacted you there.

Your questions it is not completely clear to me.

1) What is the minimum and maximum diameter of your gear and how did you measured it?

2) Are there any sharp 'concave' corner in your part? If yes, you cannot produce them by any EDM without any 'trick', shall we say. On the wire EDM in particular you will always have as a limiting radius of your round the diameter of your wire. So the thinner the wire the 'sharper the corner'.

3) What an 'involute shape/line' and 'geometrical integrity' are is not clear to me.

4) I have never used your machine. However if it as described at http://www.mikrotools.com/apps/technology.php , it seems not too bad (EDM and ECM on the same machine looks very promissing).

5) Minimising the surface roughness should be achieved by reducing the ammount of material that you remove in each single spark. To achieve this, it said that you should have to deliver as less energy as possible to your part. You can achieve this by setting the lowest discharge voltage, discharge current and discharge time setting on your machine. Now, how these quantities are mapped to actual set-up parameters on the machine depends on the machine and should be stated somewhere on the manuals of your machine, although I have never found any clear machine manual. Needless to say, the control algorithm of azis movement of the machine has a germane role in determining the variaty of different pulses/sparks that you might have during a machining operation. The polarity also should play a part. I would bet that you are using the wire as cathode (i.e. minus, i.e. reverse polarity), because you have a wire-EDM and you are aiming at micro feature furhtermore. But is this better than using it as anode from a roughness point of view? It is arguable. It should be better from the point of view of the wear though.

As you can argue, this condition of minimum roughness hinders the material removal rate (MRR). Should you want high MRR, you would get higher surface roughness. Some trade-off I would expect to be needed.

But why not to exploit fully your wonderful machine by performing roughing operation by EDM and a subsequent finishing operation by ECM (by definition no heat affected zone, no recast layer, no residual tensile stresses at all without any doubt that might persist in micro-EDM)? You cannot make the tool, eh? Too many partial differential equations involved in the design, eh? And then, even if you design this ECM tool, how to make it? With EDM again? So you use EDM for maching a tool that you use with ECM for finishing an EDM part. Confusing, eh? A kind of 'meta-EDM' tool :). But what about using an 'ECM-milling' strategy? I am going too far.

If you want to discuss it further or provide more details, feel free to contact me. I don't check this site on a regular basis (just not to say never).

I hope this might help.

Kind regards,

Carlo

"Multi famam, conscientiam, pauci verentur." (Many fear their reputation, few their conscience. -Pliny, Letters)