Projects for Senior ME's, are these too difficult?

[Q_Goest]

A local university mechanical engineering department has a program for undergraduates in which the students are given a "real world" project. The project is funded by local industry, and the company I work for is interested in funding such a project. The project costs our company $1500 with the students doing all the work. I have a few ideas I'd like to get funded, and would like some feedback from the students here, tell me if you think either of these projects would be too complex for senior ME students from a well rated engineering college.

Project #1: Computer simulation of a cryogenic pump
Abstract: Students will create a computer program which simulates the various phenomena occurring inside a reciprocating cryogenic pump. The program may be written using any platform capable of being run on a desktop PC using Windows 2000 or Windows XP. The end result should be user friendly and intuitive to use and be capable of producing various outputs such as graphs which can be viewed using Microsoft Excel.

Discussion: The program must simulate a reciprocating cryogenic pump, consisting of a suction valve, discharge valve, and reciprocating piston. A simplified diagram is shown in Figure 1.

(Note: The figure I have is too large to upload, check this link for a similar figure.) http://hplc.chem.shu.edu/NEW/HPLC_Book/Instrumentation/pump_hed.gif

The pump shown in Figure 1 draws fluid into the head (or compression chamber) through the inlet check valve during the suction stroke. As the piston moves from the top dead center position, the volume inside the head increases which causes the pressure in the head to drop. This pressure drop results in a force on the inlet check valve, which is opposed by the spring load on the valve. As the pressure load overcomes the spring load, the valve accelerates to the open position depending on the forces on the valve and its mass. Another phenomena which occurs during the suction stroke is that some fluid may boil if the pressure drops below its saturation pressure. The pressure inside the head during the suction stroke is also affected by the movement of the poppet into the head which takes up volume. During the beginning of the stroke, this is especially important, since the amount of volume displaced by the poppet is significant in relationship to the increase in volume caused by the movement of the piston.

As the cycle continues for the full length of the suction stroke, the poppet will hit a stop preventing it from opening further. During the time the poppet is opening, the pressure drop across the valve will be a function of the flow rate through the valve. Also during the suction stroke, heat will be entering the compression chamber and thus the fluid. This heat may boil some of the fluid during the suction stroke, just as some of the fluid may boil due to a drop in pressure across the inlet valve. This heat flux will be a direct input, and will not need to be calculated using heat transfer equations, only the thermodynamic state of the fluid will require calculation. As the piston comes to back dead center, the spring load on the inlet valve will cause the valve to close. Again, the closing forces operate on the valve mass, accelerating the valve to the closed position.

The compression stroke is very similar to the suction stroke. As pressure increases, some fluid will condense and pressure will rise. Eventually, the fluid pressure will increase enough to force the discharge poppet open. As this happens, the spring load and mass will affect the poppet's acceleration, along with the pressure drop across the valve which is dependant on how far the valve has opened.

End Project #1

Project #2: I haven't written up yet, but essentially it would be to come up with a similar computer program to model an eductor. The eductor would only be for incompressible fluids, such as water or oil. The input for the program would include critical geometry such as nozzle diameters, throat diameters, diffuser geometry, etc, along with input and output pressures. The output would be fluid flow rates in and out. This one seems much easier to me.

Very little public research is available on these, #2 having slightly more. The idea would be to use basic principals to come up with these programs.

So do you think either of these projects would be too difficult for engineering students? If they are too difficult, what do you think could be done to make them easier? Also, any questions on the projects might help to head off the student's confusion as to what might be expected.

 

[enigma]

I think that writing the program to do the fluid dynamics _from scratch_ would be much too difficult, unless you can deal with all sorts of simplifying assumptions.

Modeling a complex machine in an already developed software platform like Pro-Mechanica or Femlab would probably be a more reasonable solution.

IMO Femlab would probably be the better tool to learn and use. If needed, you can code your own differential equations if the pre-existing modeling tools don't cut it for the simulation.

 

[Q_Goest]

Thanks for the feedback. Those programs certainly aren't ruled out, though I've not used them personally I suspect they're easy enough to learn.

How do you feel about using Excel for something like this? How familiar do you think students are with Excel?

 

[Q_Goest]

The reason I ask about Excel is that is what we typically use for most calculations/programs here. We have a fluid properties database which allows you to bring properties into your calculation, properties such as density, fluid enthalpy, entropy, etc... So all of the properties one needs to do the calculations are easy enough to bring in using a call in Excel.

 

[enigma]

I've seen some amazing things done with excel. I personally only use it if I need to see what happens to a complex system when one or more variables are changed (used it for preliminary sizing of a rocket for one of my projects). Most things though, I code in Matlab because it's more robust than Excel. I also loathe Excel's plots and graphs.

Femlab is not easy to learn. I've played around with it some, and it has a learning curve like a brick wall. Still, if you become good at it, you are practically guaranteed a job. The results of some projects I've seen which have used it are simply amazing, too.

 

[Q_Goest]

Thanks again for the thoughts, Enigma.

Anyone else? What will make this project easier? Assuming you had this project to do, what would you ask to get started?

 

[cronxeh]

well if I was to do this project, Id get 2 ME's and one ChemE in a well ventilated room, a box of mushroom pizza and a jug of coffee. if you want us to code the whole thing from scratch id tell you to look for someone else. but if you want this thing done we'll be using MATLAB and some cfd software. it could take a week or two

 

[maglifter]

Cronxeh, you really have to make your LPG powered airplane first, before you give advice to anyone else. That was the deal, remember?

 

[cronxeh]

im still not done reminiscing over the "tiny little tourist blimp".

 

[russ_watters]

well if I was to do this project, Id get 2 ME's and one ChemE in a well ventilated room, a box of mushroom pizza and a jug of coffee. if you want us to code the whole thing from scratch id tell you to look for someone else. but if you want this thing done we'll be using MATLAB and some cfd software. it could take a week or two

This leads to my question: these sample projects seem difficult, but not very big. How big is this project supposed to be? Many senior design projects are designed to take a full-year. However, a significant amount of that time is spent doing proposals, writing the report, and preparing the presentation. If the school year is 9 months, the project should be expected to take about 6 months, with 3 people working up to 8 hours a week each on it.

 

[Q_Goest]

Thanks Cronxeh! So this is an all-nighter? Hmmm... Maybe it'll need to be expanded!

You ask a good question, russ. These are 9 month projects. I think what will take time is developing the understanding of how these phenomena can be modeled. What equations to use and how to use them when the variables keep changing. I've done similar projects before with students, and generally the way it goes is we set up some regular meetings, maybe a few weeks apart or so, and I'll help guide their work. In this case, the projects might be very difficult but as you say, they may not take so long to actually code once you get an understanding of everything.

In the case of the recip pump, the suction and discharge valve have to be modeled using fundamental principals (ie: F=ma) but then the flow through those valves is dependant on how far open they are, and how fast they actuate. The valve flow equations aren't given in college courses.

The fluid in the pump undergoes some pressure changes, so changes in state need to be calculated. The entire thing needs to be done in an iterative fashion, you start with a point in time, apply the equations, and see what happens after some very small time interval where the piston or valve may only move a few thousanths of an inch.

This actually lends itself fairly well to Excel because you can do a set of calculations which model everything over the very small time span on a single row, then copy and paste that row down, making hundreds of rows. Some if/then statements in each block limit such things as valve movement or piston motion.

So does that sound difficult, or time consuming or both? I think explaining it to folks here helps me in understanding what the students are going to want to know, so thanks for asking.

 

[maglifter]

im still not done reminiscing over the "tiny little tourist blimp".

Cronxeh, let's settle this once and for all by making a tiny little LPG powered tourist blimp together. What do you say?

 

[cronxeh]

that actually sounds very doable and reasonable.

 

[Q_Goest]

Thanks again for the feedback.