DONE

I'm all done with my thesis now. Wow. Ok, that's all. No more stress from THAT. whew... Quite a journey.

Abstract 4

This study aims to get close to the tension-tension portion of the Forming Limit Curve (FLC) of 38 um thick CP Gr2 Titanium with the use of the bulge test. With the miniaturization trend of technology, the forming abilities of Ti have to be evaluated because it is well known that the formability changes when moving from macro to micro scale. The hydraulic bulge test creates pure biaxial tension by clamping a flat foil sample to obtain a fixed boundary condition and then applying pressure on one side to promote material deformation. In order to ensure that pure biaxial tension has taken place, the strain history will be recorded by taking measurements of the strain at specific pressure intervals from the start up to the burst pressure. If the strain history has equal major and minor strains, pure biaxial tension is confirmed. Strain is measured by studying the deformation of a grid of 50 um in diameter circles that are on top of the thin foil. This thesis tests four different bulge diameters of 20 mm, 15 mm, 10 mm, and 2 mm. When bulge diameter decreases, size affects are expected to increase. Theoretical calculations will be compared to LS-Dyna simulations and experimental results to check whether the theory continues to apply when moving to the smaller diameter thin foil. The wide range of diameters tested from 20 mm to 2 mm will allow for clarification of any trends when undergoing miniaturization in forming.

Abstract 3

Title:

Hydraulic Bulge Testing of CP Gr2 Ti to Quantify Size Effects on the Forming Limit Curve when Undergoing Miniaturization

Abstract Draft 3:

This study aims to get close to the tension-tension portion of the Forming Limit Curve (FLC) of 38 um thick CP Gr2 Titanium with the use of the bulge test. The FLC helps to predict the forming behavior of sheet metal by showing safe and failure strain zones with minor and major strains as the axes. Various tests, including the tensile test, limited dome height test, cruciform test, and bulge test are used for obtaining data for different portions of the FLC. With the miniaturization trend of technology, the forming abilities of Ti have to be evaluated because it is well known that the formability changes when moving from macro to micro scale.

The hydraulic bulge test involves clamping a flat foil sample to obtain a fixed boundary condition and then applying pressure on one side to promote material deformation. If the boundary is circular and secure, pure biaxial tension takes place at the top of the bulge as the material deforms. Micro limited dome height tests already produce excellent FLCs, but the equibiaxial portion is difficult to obtain due to frictional effects. The bulge test is essentially friction free due to the use of hydraulic pressure, and its data can be used to complete the FLC for CP Gr2 Ti. In order to ensure that pure biaxial tension has taken place, the strain history will be recorded by taking measurements of the strain at specific pressure intervals from the start up to the burst pressure. If the strain history has equal major and minor strains, pure biaxial tension is confirmed.

Strain is measured by studying the deformation of a grid of 50 um in diameter circles that are on top of the thin foil. Since the Ti is 38 um thick, the strain at the top of the specimen is assumed to be approximately the same throughout the thickness. A scanning electron microscope takes pictures of the strain zone of interest. After pictures are taken, ImageJ software is used to fit ellipses to the deformed circles. The major and minor axes of the fitted ellipses are used to calculate the major and minor strains. 

This thesis tests four different bulge diameters of 20 mm, 15 mm, 10 mm, and 2 mm. When bulge diameter decreases, size affects are expected to increase. Theoretical equations for the bulge test exist, which include parameters such as pressure, thickness, dome height, diameter, radius of curvature, and material constants. Theoretical calculations will be compared to LS-Dyna simulations and experimental results to check whether the theory continues to apply when moving to the smaller diameter thin foil. The wide range of diameters tested from 20 mm to 2 mm will allow for clarification of any trends when undergoing miniaturization in forming. This study will compare theoretical, numerical, and experimental results in order to bridge the gap between the macro and micro scale.

Abstract 2

This was my first rough draft of the abstract. Bad.

Commercially Pure Grade 2 Titanium is commonly used in chemical, medical, and aerospace industries because of its high specific strength and excellent corrosion resistance. In order to limit the cost and weight in design, an accurate forming limit diagram (FLD) is desired to know the limits of the 38 micron thin foil's formability. In order to determine the equibiaxial tension portion of the FLD, experimental testing will take place with the hydraulic bulge test. In order to see any affects of bulge diameter, three different tool dies of 10 mm, 15 mm, and 20 mm will be utilized. The major and minor strains will be determined by first capturing images of the post-test grid with SEM and then using Fiji/ImagJ software. LS-DYNA software will also allow for a comparison between computational and experimental results. A linear 45 degree strain path is expected for each die diameter bulge test. A decrease in diameter is known to increase the burst pressure. The resulting forming limit diagram will help with efficient designs using Titanium thin foil. In the future, the affects of a combination of strain paths could be analyzed to determine potential new manufacturing methods.
I guess my idea of an abstract is different from what is needed for a thesis proposal abstract, which is ok. Now I will make another attempt. Some new aspects:
-Explain bulge test
-Add reasoning: testing to see if macro-scale equations still apply when going to smaller scale
-Clean up grammar, avoid using the same phrases multiple times


It is well known that the formability of metals change when moving from macro to micro scale. This study aims to get close to the tension-tension portion of the Forming Limit Curve (FLC) of 38 um thick CP Gr2 Titanium. The hydraulic bulge test involves clamping a sample to obtain a fixed boundary condition and then applying pressure on one side to promote material deformation. If the boundary is circular and secure, pure biaxial tension takes place at the top of the bulge as the material deforms. Micro limited dome height tests already produce excellent FLCs, but the equibiaxial portion is difficult to obtain due to frictional effects. The bulge test is essentially friction free due to the use of hydraulic pressure, and its data can be used to complete the FLC for CP Gr2 Ti. In order to ensure that pure biaxial tension has taken place, the strain history will be recorded by taking measurements of the strain at specific pressure intervals from the start up to the burst pressure. A grid of 50 um in diameter circles are placed on the foil and are measured post-deformation with the use of SEM and ImagJ. This thesis tests three different bulge diameters of 20 mm, 15 mm, and 10 mm. When bulge diameter decreases, size affects are expected to increase. Theoretical calculations of dome height will be compared to LS-Dyna simulations and experimental results to check whether the equations continue to apply when moving to the smaller diameter thin foil. This study will compare theoretical, numerical, and computational results in order to bridge the gap between the macro and micro scale.


I'll stop it there for today.

Title and Abstract

I have my committee together, and the next step is to propose my thesis. I need to have my ideas solidly together and be prepared for any questions.

So first: what will my title be?

Hydraulic Bulge Test of Commercially Pure Grade 2 Titanium for the Forming Limit Diagram on the Mesoscale

Hydraulic Bulge Test of CP Gr2 Titanium for Right Hand Side of the Forming Limit Diagram

Forming Limit Diagram of CP Gr2 Titanium with Hydraulic Bulge Test

Something like that? I should emphasize that I am focusing on the right hand side and particularly equibiaxial tension. Ok, I can work with that.


Abstract
-Intro: Why is this research important? (1-2 sentences)
Commercially Pure Grade 2 Titanium is commonly used in chemical, medical, and aerospace industries because of its high specific strength and excellent corrosion resistance. In order to limit the cost and weight in design, an accurate forming limit diagram (FLD) is desired to know the limits of the 38 micron thin foil's formability.

-Methodology (1-3 sentences): specific approach (theoretical, experimental, computational), measured variables and control parameters, (not step by step)
In order to determine the equibiaxial tension portion of the FLD, experimental testing will take place with the hydraulic bulge test. In order to see any affects of bulge diameter, three different tool dies of 10 mm, 15 mm, and 20 mm will be utilized. The major and minor strains will be determined by first capturing images of the post-test grid with SEM and then using Fiji/ImagJ software. LS-DYNA software will also allow for a comparison between computational and experimental results.

-Results (3-8 sentences): In my case, expected results. Specifics
A linear 45 degree strain path is expected for each die diameter bulge test. A decrease in diameter increases the burst pressure.

-Conclusion (1-2 sentences): significance of results, future steps
The resulting forming limit diagram will help with efficient designs using Titanium thin foil. In the future the affects of a combination of strain paths could be analyzed to determine potential new manufacturing methods.

And now, all together:
Commercially Pure Grade 2 Titanium is commonly used in chemical, medical, and aerospace industries because of its high specific strength and excellent corrosion resistance. In order to limit the cost and weight in design, an accurate forming limit diagram (FLD) is desired to know the limits of the 38 micron thin foil's formability. In order to determine the equibiaxial tension portion of the FLD, experimental testing will take place with the hydraulic bulge test. In order to see any affects of bulge diameter, three different tool dies of 10 mm, 15 mm, and 20 mm will be utilized. The major and minor strains will be determined by first capturing images of the post-test grid with SEM and then using Fiji/ImagJ software. LS-DYNA software will also allow for a comparison between computational and experimental results. A linear 45 degree strain path is expected for each die diameter bulge test. A decrease in diameter is known to increase the burst pressure. The resulting forming limit diagram will help with efficient designs using Titanium thin foil. In the future, the affects of a combination of strain paths could be analyzed to determine potential new manufacturing methods.


After meeting with my advisor, the above abstract is too "short." I need to make it sound more technical. Ehh, why? I like making my writing easy to understand. But oh well.

So I'm working on organizing myself. I have various projects and duties to attend to. I want to produce a good, actually beneficial thesis, and I can do it - through the use of organization. Perhaps updating this blog more can help. I have taken my final for my class already, and all I have left in front of me is a design project for an interview, a lot of grading, and this thesis.

Update 4/18

It's crunch time, and I'm not talking about lunch. The bulge test has to give me what I need in terms of a thesis. I will need to propose a thesis real soon.

What do I know already?

CP Ti grade 2 is the material I am working with. That is, commercially pure titanium grade 2. This material is used a lot in industry. It has superior corrosion resistance and has a high strength to weight ratio. This means it can supply good performance with a lesser amount of material. Because of its corrosion resistance, it is applicable in chemical, medical, and aerospace industries. It is quite expensive though, so that limits its reach into other industries that can easily use other cheaper material.

The forming limit diagram is helpful in determining how much stretching can take place before the material fractures. It plots the major true strain vs minor true strain against each other. What does this mean? Well, here's a picture:

From this picture, you can see some cool things. The darker red line is what people want to achieve with testing. Uniaxial tension is achieved with a tensile test. It's where the material is stretched in one direction and shrinks in the other direction. Poisson's ratio is involved in there too. The left side of the FLD for CP Ti gr2 has been tested for already at NIU. Micro limited dome height was used to get that data. However, the right side is more difficult to get with that method. There is friction involved with the limited dome height test. So here is where my thesis comes in. 

The bulge test uses nothing but air to deform the material. That means that there is essentially no friction involved, which is good for data. For a perfect circular bulge, there is equibiaxial tension. Both the major and minor axes are strained by the same amount. Perfect! This can find the tip of the right hand side FLD. My thesis is essentially just trying to find that right hand side.

Is this enough of a thesis? I don't think so. For this reason, I will try to use different diameters of bulge to find out if there is an effect on the strain path. Perhaps the smaller diameter bulge tests behave slightly differently. Who knows? Furthermore, I can take pictures at different points of the bulging process. This can give me a strain history at various pressures. Maybe this is important in some way?

This has been studied on the macro scale before, but we are working with 38 micron thick material here. This means that the macro scale experiments may no longer be an accurate model for our purposes. It's time to do some more literature review and ask some questions.

Some questions:

1. What macro scale experiments have been conducted already on this material?

2. What experiments of similar material have been conducted already?

3. If other material has been tested, what can be expected to be the same or different?

4. What constitutes a macro vs micro size? 

5. Have other materials shown variations in macro vs micro characteristics?

Update 4/6

Oh boy, do I have a bunch to talk about! I kind of forgot about this blog, honestly. I've been able to gain a lot of traction with the bulge test, and I think that's the way that I'll end up going with my thesis. Yes, the biaxial tensile test machine is still in the works, but I have actually performed bulge test experiments already.

So where shall I go with this? I ended up getting a new MultiMoto shield for the arduino. It worked flawlessly, so I am now able to control the speeds of the four linear actuators independently. All I would need to take care of with that is to make speed vs PWM input curves for each actuator. That way the speeds could be perfectly controlled as desired. The other MultiMoto shield that only half worked (before spring break) was returned to the seller. They tested it and determined that it was indeed a dud! Yay, it wasn't my fault that the older shield didn't work! Hallelujah.

For the biaxial machine, though, we still need some grippers to hold the 38 micron titanium specimen. Yes, that may be difficult. The grippers themselves aren't too difficult to make, but aligning them is a challenge in and of itself. The current clamping method is quite rickety at the moment.

But Hallelujah, the bulge test is where the fun has been happening! Imagine bolting some metal together with some o-rings and just pumping in nitrogen to 400 psi. Yeah, it's quite nerve-racking. But after a few times you find out that it's actually not that bad of a pop when a failure occurs.

Anywayyyyy, I got the machined parts back from the machine shop and was super pumped when my o-ring grooves fit the o-rings perfectly. I love how quickly the machine shop gets stuff done, too. It's definitely a change of pace from some of my past experiences.

Above you can see the beautifully shiny 303 stainless steel parts. Aren't they pretty? <3 Those o-ring grooves are meant to hold the nitrogen in the system, and they sure have been doing a good job. Hurray for the random internet source I used! 

Since we were really scared and didn't know quite what to expect, I clamped the bulge test parts underneath a table and put a blast shield in the way. This reminds me a lot about my senior design project where we just spun a bunch of magnets at a couple thousand RPM. I should consider safety a bit more in the future maybe. In the picture above, though, you can clearly see some bulging occurring in the titanium sheet.

And above here you can see some bent metal. Except instead of titanium it was aluminum. This aluminum burst at pressures of maybe 100 psi or less. Very weak! Because the pressure was still so low, the metal didn't fly off anywhere. That's what I need to happen with the titanium samples. However, that will be quite unlikely. Therefore, a different plan of attack is necessary.

Here you can see a picture of a bulge test of the titanium at 430 psi. The 440 psi test burst, so there aren't any good pictures of that. Notice the crease around the edge of the 20 mm diameter bulge? Yeah, that's the sign of a stress concentration (more on that later).

Here is a picture of all of the samples we tested. Titanium bursts at around 440 psi while Aluminum burst at quite low pressures. The 2 mm bulge test on the bottom left of the picture did not show much promise at all for us. We simply cannot achieve enough pressure to bulge that small diameter.

We tried using strips of Ti over whole blank sizes of other material. 100 micron thick Al did not work well. Two 50 micron blanks of brass was too strong. A single 50 micron brass blank underneath a strip of Ti seemed to work well. By using different strip geometries, different strain paths can be achieved throughout the testing. Also notice that the Ti on the bottommost sample still failed at the edge of the bulge. This indicates a stress concentration. I have had the inside radius of the tooling cut down to a 1.5 mm radius. This has worked well thus far.

More later.

Update 3/1/17

All righty. I got the power supply working perfectly, and I even got the DPDT switch working correctly. Now, however, I am backtracking a bit. Yes, this system is able to control the directions of the actuators, but it has no way of controlling the speed. In our application we have no use for a machine that pulls apart at 0.4 in/s. Nope. Therefore, I need to get my hands on another MultiMoto shield and get it working with an Arduino. That's all that's needed, man.

For the bulge test stuff, I talked with the machine shop and got a general feel for what they can and cannot do. My drawings have been in metric units, but they work entirely in imperial units. Darnnnn. After looking up pricing of parts, I understand why America is staying with imperial units instead of switching over with the rest of the world. Things are just cheaper! Therefore, I will modify my dimensions and part decisions to go with the cheaper options.

I feel confident that all of this will end up fine. Steady work, no rush.

Bulge test dimensioning

The bulge test parts must be made. The idea for this is that a nitrogen tank will supply nitrogen at incremental pressures to a thin foil until it bulges and breaks. In order for this to work, a fixture is needed to hold the foil in place and apply pressure.

One aspect of this test includes o-rings. Page 109 of this reference gives me some baseline to go off of for my o-ring groove dimensions. I'll be using a 2 mm thick o-ring, just because it looks like a good size. 

Feel the power

My power supply came in the mail, yay! I ended up buying this one. If it performs as advertised, it should give us 12V up to 30A. Cool! It should work well. I still need to figure out how to wire up everything though. I have a wall plug that I cut open, but I haven't figured out which wire should go into which port.

This website might help determine what each wire should be attached to (wall to power supply). (EDIT: The European wiring worked! Green/yellow ground, blue neutral, brown line)

How to wire a DPDT instructables looks good enough in teaching me how to wire up the DPDT switch I bought last week. By the looks of it, I may need to do a little bit more shopping for some electrical connections and possibly a crimping tool. All I need is to get this thing moving. After it moves, we may run into more issues. (Yay)

If you are reading this, you may have by now figured out that I am just using this as a place to drop a bunch of links for my own future reference. My plan is working out perfectly! hahahahahaha

Switching things up

Frustration was the word of the day yesterday. I guess frustration stemming from inexperience. The motor shield appeared to be in place to be usable, and it actually worked for a little bit of the time, but  that was until it all kind of went wrong.

Some of my soldering came loose on the big motor/power attachment pins, which prompted a re-solder. I'm not the one who performed the re-solder, and I have a feeling like that might have messed things up a bit. I tried hooking up the power to the shield again, and the current was quite high. The voltage was low because of the high current draw. Yeah, things did not end well... I knew that because of the large spark that was momentarily about the size of a quarter. Gah. Frustrating.

While I am able to get the Pololu shield working, it requires very careful soldering. Long story short, I am not willing to put that level of care into it at this point in time. We are going to buy a power supply and connect the motors with some switches.

Three switches will be used: one to switch on/off the x-axis, one to switch on/off the y-axis, and one as a master switch. The master switch is a DPDT center-off switch. This allows for both reverse and forward direction. The only concern is that we won't really be able to control the speed of the actuators. Maybe a rheostat could be used for that?

So yeah, that's where we are with that situation. I have wire and three switches from RadioShack. The store ended up being in a "going-out-of-business" state, so I got all three switches for around $6. Sweet!

Each switch is rated for 12 Volts and 25 Amps (or something like that). Now I just need a power supply that can give us exactly what we need. 12 Volts. From what I understand, a power supply amperage rating does not indicate that it will actually give max power all the time. If it is rated for 20A, it has the ability to supply 20A of current, but will only supply whatever amount the load requires.

Soldering the Pololu shield

Today has absolutely flown by. After figuring out how to transport my roommates soldering stuff to the engineering building, I had to get my confidence sufficiently up to complete the job. So what did I do? I just kind of jumped into it.

Figure 1: Soldering setup

I had a nice little setup in the lab, as can be seen in Figure 1. A slightly wet sponge, a heating up soldering iron, and shaky hands.

Figure 2: not the prettiest job

I'll be the first to admit that my solder job doesn't look the best. I tried this alignment technique using the Arduino pins themselves, but I think the pins are misaligned by a little bit, so my solder job went a bit misaligned. That is just for two rows of pins though, as seen in Figure 2. Two other rows came out totally fine. Even the jumper pins and motor/power junctions came out decently well.

Figure 3: Oh babyyyy

Figure 3 shows the underside of this motor shield. It's the result of my first time soldering. Happy days! Next up I'll have to find out if it actually works as it's supposed to work. I really hope it does. Maybe I'll hook up a low power DC motor I have first to make sure it actually works, and then I'll hook up the big-boy actuators.

I hope I didn't burn anything..

Here are a few more pics of the shield attached to the Arduino.

Psyching myself up to solder

Oh boy, the Pololu motor shield came in! It came in yesterday, actually, and it's smaller than I thought it would be. I'm going to have to learn how to solder now. I mean, I could probably do it ok, but I need to make sure that I don't mess anything up. I want it to be operational as soon as possible.

So with that being said, I've been watching some videos that involve soldering or teach how to solder.

Video 1 - A shield being soldered for Arduino

Video 1 shows the first video that I've watched. It's just a guy soldering stuff. I'm anticipating mine to work a bit differently, though, because I have pins that have inserts for 22 AWG wire at the top. Hmm.

Video 2 - Soldering for beginners

Video 2 gave me some good hints on how to solder. Since I'll be soldering small components on a board, I will have to keep the temperature on the low side (600 degrees Fahrenheit or 315 degrees Celsius).

Steps I've noticed:

1. Clean off connections and soldering iron.
1. A wet sponge can help for cleaning off the iron.
2. Apply some solder to the tip of the iron first.
3. Heat up the metal that you want to solder.
4. Apply solder to the heated metal (and not the iron tip).
1. This is because there is flux in the solder which helps make the connection.
5. Quickly pull up the iron to get that nice bubble look
6. Don't heat the board for too long.

I think I can do this! I'm just somewhat worried about the size of the soldering iron tip. My roommate has an iron, but the tip is quite large. I have an iron too, but it has no way to control temperature. I can find out if the engineering building here has an iron for me to use. 

Video 3 - More soldering!

Ok, by the ending of Video 3, I feel pretty good about the steps. I should practice on some cheaper items first, though. Cool cool. I will tackle this issue tomorrow when I have free time. 

Specs: Arduino motor shields

It's all fine and dandy that we have some linear actuators in our possession, but moving them is another issue. We need to control the speed at which they move outward, along with some sort of initialization logic. The actuators need to all start with the grip ends at the needed locations, followed by being retracted, causing the tension that we are looking for.

In order to control the movement, we are using an Arduino Uno with a motor shield. The particular motor shield that we are hoping to use is the Robot Power MultiMoto. It appears to be perfect for our needs. 

Figure 1 - Robot Power MultiMoto

Some specs of the MultiMoto (seen in Figure 1):

• 6.0V-36V battery voltage
• Four independently controlled channels
• 6.5A continuous current on each channel (8A max)
• Nearly blow proof

Wow, great! Our actuators have a limit of 6.5A as well. We need to give 12V to each actuator, so that's covered here too. 

But sometimes things are too good to be true.. We tried it out and something went wrong. I wasn't actually there for this, but the four actuators were all moving ok, one of the wires came out (or something like that), and it was powered down. Somehow the Arduino was fried. The MultiMoto doesn't seem to be functioning either. So we are in process of replacing the MultiMoto.

In the meantime, I am trying to find something a bit cheaper to buy so I can at least power two of the actuators. If I can get the machine working in one axis, it would be easy enough to replicate in a second axis.

The requirements for my personal motor shield purchase are:

• Able to handle 12V
• Power two motors
• Able to handle at least 3A per channel

Figure 2 - Arduino Motor Shield R3

I found these handy dandy instructions to help set up an Arduino Motor Shield R3 (as seen in Figure 2) (http://www.instructables.com/id/Arduino-Motor-Shield-Tutorial/?ALLSTEPS). 

It appears that this shield is really only able to handle 2A per channel (for two motors), or 4A for a single channel. Hmmm. Will that be enough? I guess if anything, I can control on actuator. It will set me back about $25 (amazon link).

Figure 3 - Pololu Motor Driver shield

But I shouldn't settle for less than what I need. Figure 3 shows another shield I found. It's the Pololu MC33926 dual motor driver shield for Arduino. Again, it'll set me back $36. But it seems to have the specs that I'm looking for as a short term replacement. 5V-28V operating voltage. 3A continuous output current per motor (5A peak). It would just require a little bit of soldering, which I'm sure I could do. Pololu motor shield manual

I feel like I should spend an extra $11 to be able to power two actuators sufficiently instead of just one. If this ends up working well and the replacement MultiMoto still proves problematic, we could incorporate two of the Pololu's.

You may have noticed I've posted links to Amazon Prime sources. This is because I have Prime for a month and I'm trying to take advantage of it! If I can get these parts in a day, it would really help me have some more time for finishing it up too. I don't really want to deal with long shipping times. I have work to do.