 Okay, good afternoon. My name is Ramesh Singh and I'm a professor in mechanical department and I work in a lab called machine tools lab So today what we'll do is we'll talk about what I do as research in a machine tools lab And there are a lot of products that have come out of my lab And I'll actually take one product in a particular laser vent cleaning and walk you through the entire Development process of that product. So first thing is that we are all at ID Bombay And then I would just like to give you some details about ID Bombay We have 15 departments, 16 centers, 4 interdisciplinary programs, about 650 faculty members A decent amount of R&D funding 10,000 students give or take out of which 30% of our student body is PhD So despite what we think that it's a UG centering institution We have a significantly high number of PhD students and a few of them work in my lab Who are involved with these kind of product development and and basic research which leads to product development And then we have support staff and infrastructure which actually supports our research And I am part of mechanical department. Mechanical department does a lot of things So the key areas of research is something called micro nanotechnology computational methods, fluid mechanics and thermal sciences Manufacturing then there is bioengineering basically that's the need of the hour because lot of the products of future Related to healthcare and also to enhance our quality of life And then solar power energy because energy again is a big issue So we want renewable, so solar energy and solar power is also one of the focus areas of research So basic research area which I do in my lab is first one is micromachining So either I can subtract material which will be machining Machining are of different forms. I can do tool based machining. I can do electrochemical machining I can do electrical discharge machining. I can use different things to remove material Or what I can do is I can add material which is which we do in additive manufacturing Or I can do joining kind of a process where we can do well multiple things and make a product out of it Or I can do something like casting and forging which is mass conservation process We don't try to develop the process. We also develop the machines to work with it Because these machines are commercially some of them are available Very few though, but they're very expensive So I'm innovative and I actually go ahead and build those machines And then once you build the machine there are some challenges the process itself Because the tools which I use just to give you some idea what tools I use in machining can be one fourth of your human hair 20 microns The second thing what I do is I call it something called flexible reconfigurable fiber laser manufacturing The reason I use that is that I don't have money to buy five different machines So what I do is I buy one laser and I and I basically make changes to the The spot size the powers So basically I can do very high energy intensity To ablate the material sometimes low to heat the material so I can I can melt the material I can heat the material I can vaporize the material using the same laser by changing the optics And a combination of other things to do certain things should do like texturing sometimes hardening sometimes Adding material and printing. So basically printing is nothing but you add powder and melt it and create layer by layer a structure So that's what you do. So I actually do all those things with with lasers Then Since I teach mechanical engineering I have to Understand what is happening mechanics wise Since I'm doing all these things these days is everybody's doing bio, right? So I said I cut steels. Why don't I cut tissue? So a couple of students what they do is they try to figure out That if I take a tool and poke a tissue What will happen? Earlier days people used to just do a big incision and do the surgery nowadays You can come out of surgery in a day because they do a very small thing and then they would do something called laparoscopic surgery They put a cannula and then they'll put a tool which would go ahead and do it local surgery Now they're not seeing it although they do have some cameras to look at it But they don't have full vision now a lot of it depends upon the feedback So I'm cutting lever. I'm not cutting something around it, right? So they feel it So a lot of it would be how the force response of a tissue is so we want to study that And then get that idea to make some simulators or make some tools where people have A feel a priori before he does the first surgery if he does 10 surgery, it's okay But then how do you make sure that the guy goes And does the right things because bone feels different liver feels different Stomach tissue will feel different. I have to go do liver surgery and I go cut my stomach The only way the only thing stopping me is stomach feels different If I'm poking there it feels different create more realistic surgical simulators We only have managed to just get the forces from different models So this haptic feedback would be based on this knowledge that what tissue gives what forces So this information will build the backbone of it Now these are some snapshots of the lab. We have a whole bunch of machines Since it's the name is machine tools lab. All we do is we have a lot of tools We have a timer laser Wire electrical discharge machining We have something called white-clad interferometer for surface toughest measurement So if you machine something you need to measure the roughness. That's the tool for that There's a micro machining center. It's a commercial thing and then measuring microscope We have lots of surface roughness and metrology characteristics because what we make right? I do micro machining So I should be able to measure things which I do. So a lot of these equipments actually help me measure what I do So I need to be to be able to measure few microns if I'm making a channel of 100 microns I should be able to measure it. This is called a coordinate measuring machine So what it does is any shape you make it can give you a exact measurements of it This is a very old machine which we have in our lab, but the newer ones can give you 3d models You just put the product they will take data all around it. They have they have a probe They will go and get all the data surface data and build your 3d model for it A lot of my money which I do in my research comes from industry And the industry typically if I if I take money from industry I have to give them something So this is what would be probably of more interest to you being design students That how do I go ahead? So they come with a problem, right? So then I have to interpret the problem in a way Which makes sense to me Have a path to design it Have some engineering analysis to back it up whether this will work or not And then finally design and build it and it should not be ugly looking We designed the India's first ultra high speed micro machining center That I told you that there are processes for forging, right? Where what you do is you heat the material and shape it using a forge using a die or a mold Right So what happens is after a while since you're taking hot metal and come usually compressing it, right? It can wear so basically what will happen is it will not have the correct shape after that, right? And if it has cracked eventually it can Actually crack and break So these things are fairly expensive each mold can cost you up to 30 40 lakh rupees just one mold Right, and you can just treat away so We have developed a very high fidelity model which actually takes into account if the material melts and desoldifies There are changes in the microstructure And this changes what they do is they create volume dilation and transmission to its plasticity So and there is a thermo mechanical contraction because as the material cools it will contract And this contraction will not be even because the higher temperature will contract more the lower one will Contact less so there will be a differential contraction post cooling That will create a stresses and then there are transformations in the Microstructure because you cool it very fast All of them will have a role in the residual stress For those of you know what what is the stress is it basically means the moment I cool ended everything There's no load still there will be a stress in there locked stresses What it will do is if it's tensile, right? And there's a small crack the tensile will open the crack so your fatigue life gets compromised So if I do it just by depositing not knowing what kind of stresses are evolving not understanding the process I barely do a job which will fail again in three cycles Do three more four things again it will fail So if I understand the process mechanics doing the scientific way With the scientific knowledge, I'll probably be doing something which has a much longer life So that's what I'm making a good case of doing it scientifically you can do it But if you do it scientifically it will be more sustainable Whatever technology you are developing has to be enabled by science Okay Now so what we propose to do in this is If somebody comes to me and says that we have to develop a repair for that, you know what I'll do I'll actually first create a automatic scanning system. It will go ahead and scan Compare it with the solid model that what is the actual product look like And then it will identify this is the defect. These are the defect areas Then it will go ahead And take a powdered material and then do a local 3d printing there local deposition of material there And add the process parameters which will not create undesirable digital stresses So those possible limiters will come from science And the technology is the actual product and the The system profoundly that goes along with it So that's what we propose and we call it there will be the autonomous autonomous damage detection system The material deposition system and the entire thing which will come into one package would be called restoration system Filling of deformities layer by layer does that create some problems like from the actual material which is Yes, it will So you need to know what it will do Most of the material which I make by forging It's actually I so properties are same everywhere, but if I do layer by layer the property variation will be there And then I build another machine for BARC They need very high speed bearings very very high speed bearings And these things have to run 15 years non-stop Okay, so it has to be very reliable Now they want to change the material from something what they use right now to sapphire If I have to machine sapphire what will be the biggest challenge crack is a big problem Because the material and you need something harder to cut because sapphire as it is very hard So the problem was whenever you machine will create some cracks And the only thing you can use to machine is diamond So either you buy a super duper expensive machine to do something like this Which there are very few in India So we said that okay, we'll do it cheap. I'll run it in my lab I actually machined it in my lab with some cracks in it a little bit of cracks minimum possible I actually cut at one micron depth of cut if you can imagine pushed it to the limit Still there were some cracks To take care of that we developed the polishing process And it's a very funny polishing process. I have a cavity to polish How do you think I should polish this cavity if I have a cavity to polish? Of course chemical is a good idea because then chemical dissolution would enhance the thing But even to do that chemical dissolution evenly all the way through I need to make sure that There is relative motion between the all the points here So the first thing that dawned on to me is do idly-dosa thing Make a ball that conforms to the thing and just moves Now what will be the problem if I do something like this? The corners will rotate the velocity will keep on going Smaller and smaller at and that center it will stall. So there will be no polishing in the center So what we did was we did two things one we did it 45 degree and then rotated the ball also So we did a Double movement. So initially what we did was we did something like this We did a motion a rotating a rotating ball with With what they do in the idly-dosa motion That was a very kinematically very Unsmart way of doing it then what we did was we did this and it's a very precise location We started holding the ball itself the the cavity itself So now there is non-zero velocity everywhere and build a machine like that and of course we use Something that chemically reacts to the To the surface so the combination of mechanical and chemical polishing And give the machine to to be RC One of the guys from seet quality control came to my lab and said that when they build these tires, right? So what happens is there is misalignment between those edges Once the misalignment is there regardless of what we do everything is a waste Everything is a waste And when they do the first pass of the tire building, there is a there is a drum roll And that That tire that basically the rubber piece is rolled over it out of it is fed automatically and get rolled over automatically Now what happens is If the alignment is off so they have two laser markers And the guy eyeballs it The requirement is three millimeters over 1.6 meters And he eyeballs it like this If it is not right He will take it and do it again by hand Each time he does it he could have built four tires automatically Each time he unwinds it He is losing time for four tires of manufacturing So they came up with solution they come and tell me what we'll do will take a scale and measure it I said how on earth you can measure 1.6 meters and get a 3 mm accuracy with one mm least count of it How do you even see it? So one guy will see there other guy will see there and there is 1.5 and here there is 1.5 mm I said this is crazy there is no way to work Then he says then what I'll do is I'll design a special caliper exactly the same which I want and put it there He said you'll use caliper on a rubber on a compliant material 3 mm will compress it so every time your product will be right Put a caliper it will just compress the material compress the compliant material it's a rubber right I said So I told them the only way to do it would be that you have a good vision system Take a camera measure it real time measure it very fast do it at various locations While the tire is rotating look look look look at it as every location So I said that every 10 degrees get a data And then do it so we actually built a system for them Then every 10 degrees it measures it actually takes the The laser mark identifies the edge Measures everything how how much off it is computes real time And tells the Operator it's good or bad and locks the data to their quality system. So the entire system we designed and built further Now this this went out fine. One of my old students who has a company now designed everything Now what happened was They came to us and say that we have another problem. This is what I'll talk today primarily the vent cleaning system They came to my lab. Oh, you use lasers We have a problem The problem what they defined was very simple. They said that We have these Molds, right? So tires, you know, they will take so all this tread Is actually molded you take tire And you take a heated mold Get those shapes cool it take the tire out Now what will happen is all these molds any mold will always have vents because The material has some porosity. So the gases will come out So the gas has to be To come out and these vents would be there. So these vents what happens if I do it say about Thousand odd cycles of molding these vents get clogged and if I do more if I do more of it With the clogged vents the quality will get compromised Because my air will not escape. So there'll be porosity in the product. You don't want that Right. So the way they clean it is Somebody actually takes a drill 600 micron drill And by hand Each tire has 1600 vents 1600 somebody goes and Does it You have seen a new tire, right each tire will have hair like structure That's called a spew vent spew So that hair like structure is because you have vents So you'll actually new tire you'll see everywhere even on the radial face But on the side walls you see even in the old tire If you look at the side wall there these these tires will these Spiky things will always be there even in the older tires. Those are vent residues So this is what actually goes into the vent and that's what needs to be cleaned So they will take a thing and tap tap tap tap tap tap tap they'll clean it like that But that's a process where it is inherently very dangerous tool can break If the tool breaks what will it do So it can be in the vent that's also not required and for those of you who have never done drilling If if a drill breaks In a hole while drilling taking that thing off Is a nightmare A hole and then the other thing would be that the to take out is a bigger pain Right So now that's that's one issue which is there So what happened to them the seed guys actually Gave a tire with metal embedded in it and their entire batch was rejected By Renault These guys come to my lab up. You use lasers. Can you do something? Can you design something which actually just removes all this? drilling business I said, why don't you send me the molds? Let me do some experiments We'll use lasers to see whether we can clean these using lasers or not. So that's how this project evolved Then a lot of these diamond guys come and work with me because They polish diamonds, but they never knew What they were producing So they came to me and said that You know, I have an idea I want to do digital personalization on jewelry and gemstones So I said, what do you want me to do? So he said that can you do some digital personalization every Jewelry with somebody buys is unique You can create an app where I you can have a unique code on the on the jewelry Somebody scans it the moment you scan it that unique code is linked with something on the web where You can have your love letters pictures messages blah blah blah photographs pop up So that will be a digital personalization of the jewelry the jewelry will come along with A lot of things stored. It will have a unique optical code So I said I can actually make those codes. So what I'll do is I'll machine this small small array of codes Actually, these are just bunch of arrays of holes Each array of holes one once it's scanned. It actually gives me a unique code. It's like a barcode or something But since you cannot do it on jewelry, it will look cheap. I'll actually create a slightly fancier design Micro machine of course with tools or lasers And then you can do it So the only problem with that would be that that has to be Amplified so I'll give them a magnifying glass a lens attachment with it So anybody who buys a jewelry will get a lens attachment free with it with the jewelry which which goes into your cell phone So that will be additional tool which Will be given to the guy who buys the jewelry a a magnifying lens The ideas don't come in vacuum You need solid science behind it to generate those ideas So research is the driver for those things unless until I knew about lasers I I would not be able to make this machine So I have some idea based on my my scientific background that I can Extend it to a product or extend it to a to to a more marketable idea Now everything every small component has been scaled down So the scaling down can be in various fields. It can be in biomedical defense applications jewelry electronics Molds for small features So there is huge amount of scaling down of things So traditionally, I don't know if you understand People still used to make a lot of small things People who are making chips since 60s So they were doing it the only thing the way used to do it is they will make a mask somehow and then they would Shine a laser on it or UV lamp on it and then they would etch it away. It's called it's called lithography process But then that will do only two and a half days still used All the chips are manufactured the seven nanometer technology which is there the cutting edge technology uses lithography still But they're primarily for silicon And that's a process which you don't want to scale up And it will not be used for every material. So what I say is there is a Nanoscale manufacturing, which is good. We have the technology for it Great. I have a macro world where all my machines are there. It is in between there is a micro scale where either I have to Come from top or take these processes up there. That's a gap All those lithography processes work for silicon. So if I have to use titanium steel All these metals it's not easy to do So why don't I take the macro scale processes scale it down? So processes can be milling drilling turning EDM Now what is the primary problem? The primary problem is we don't understand the science also at that level As I told you one example, right what stiffness does Right the stiffness scales non-linearly So we have to design a process in such a way that forces are very very low Any small force will break that rule So fundamentally what we said is the way to counter that would be that you go at 100,000 rpm Normal machines run at 3000 rpm The problem at going with very high speeds is For those who don't understand dynamics Certain frequencies can be excited Any misalignment if there is even out of misalignment it will be amplified omega square r if you remember that So there will be alignment Amplifications misaligned amplifications there will be some of the natural frequency will be excited And any small vibration with a very low flexibility low stiffness tool is very dangerous too little break So you have to understand the science of dynamics very very well in the machining process to be able to make it properly that is the big challenge So I can make a machine but to operate the machine to understand the process very well