INL - Grinder Water Circulation and Filtering System

From Mindworks
Jump to navigation Jump to search

Sponsor Idaho National Laboratory
Advisor Dr. Matthew Swenson
Duration Fall 2017 - Spring 2018
Mentor Alex Olson
Team Name Keep It Clean
  • Drew Fagan
  • Jerry Kahn
  • Thomas Moore
  • Troy Sanders

Idaho National Laboratory (INL) has reached out to the University of Idaho to completely re-design a water filtration system that is to be used in a hot-cell environment. Our team is working closely with the staff here at The University of Idaho along with INL to ensure that our system meets all of the specific requirements. On this page we talk about the system we designed and built, decisions we made, and we will provide as much information on our project as possible. Overall, our project was a success and we are excited to share our work with you.

Project Definition[edit]

Problem Statement[edit]

For this project, we were tasked by Idaho National Laboratories (INL) to redesign a grinder/polisher water filtration system. The main function of this system is to rinse off a grinder and polisher within a hot-cell, filter the particulate and water solution, and recirculate the clean water back on the the grinder and polisher. The current system in use by INL needs to be redesigned because it needs cleaning often, the filter is expensive and inefficient, and the pump gets exposed to particulates from the grinder and polisher. As these are our three biggest challenges, we set out to design a system that focuses on particulate centralization and filter efficiency. Solving these issues will result in a system that can work for long durations and can be easily replaceable when needed.



Project Design[edit]

Design Alternatives[edit]

Filter Alternatives[edit]

Image Description
Filter 1.jpg This was the initial design concept that we had in mind when trying to develop a new filtering system for the water circulation system. Overall this was a really great design and it had a lot of potential, but when it was first presented to our client we realized we needed to implement 2 different types of filters. A rough filter to collect larger particles, and fine filter to collect the rest of the particles as needed.
Filter 2.jpg The second design that we made seemed to be very innovative, and it proved to be more complicated than needed. Though this design met our new constraint of needing a rough and fine filter, it didn’t provide an easy option for cleaning the system out when it came time to replace or unclog the filter. As you can see these filters worked in parallel, and that would allow the flow to exit through the rough filter and then collect particles on the outside of the fine filter. Collecting particles on the outside is where the problem lies with this design. When changing out these filters we want all of the particles to be taken out as well.
Filter 3.jpg This design is our final design and overall what we will manufacture by the end of this course. This design is simple and it consists of a design similar to our initial design but with a rough filter that fits inside of the fine filter. Our only problem with it currently is that it is hard to find filters that meet our specs, and that fit inside of each other as we have displayed. The only other problem that we have had with this design is that it would be harder to replace these filters in the environment because it would take twice as long to replace both of the filters.

Tank Alternatives[edit]

Image Description
Tank 1.png Our initial thoughts with this design was that we would maintain a system similar to the current system and have the pump and filter on far sides from one another. Our intentions were also to allow the filter to drain straight into the tank in order to eliminate some tubing in our system. It also seemed to be the best option to accommodate the pump that we decided on using for the system and would put any particulate under the pump meaning cleaning would be even more simplified.
Tank 2.png This was a similar idea to the first with the idea of having a space where particulate could accumulate if it did make it into the tank. However, it was not really a good fit with the pump we were choosing since we needed a larger clearance for the pump, and if we centered the pump it would interfere with other components to the system.
Tank 3.png This was an idea that I came up with after the first snapshot day because some students suggested a spherical tank design. The half cylinder would also make it so any particulate in the tank would go the bottom of the curve to make cleaning simple. In order to meet the volume goal we had set it would require this tank to be pretty long and if we wanted to make it any deeper we would have to add a section of straight wall. Which made this design undesirable for us.
Tank 4.png Like the half cylinder the idea for this half pill shape was a result of the first snapshot day. Just like the half cylinder it would cause any particulate in the tank to accumulate at the bottom of the curve. However, the addition of the quarter spheres on the ends would eliminate the edge at the end for particulate to get stuck in. For this design to meet the length constraints it would require a section of straight walls before the half pill though since it would basically be the full length of the system otherwise. Meaning this was a less than ideal design for our system.

Pump Alternatives[edit]

Image Description
Pump 1.png This pump from Bell and Gossett was the first pump that we considered. The head produced from this pump is satisfactory for the desired volumetric flow rate. However, our team was concerned that the materials of this pump would not be able to hand the environment of the hot cell. In addition, mounting and tubing integration would be a significant challenge with this pump. As a result, we disqualified this pump as an option for our system.
Pump 2.png This was the ideal pump given the head and volumetric flow rate requirements. At 1/5 gallons per minute we would receive 6 feet of head. In addition, all the materials are confirmed to withstand the radioactive environment. The downside to this pump is that the discharge port is only capable of a slip-on connection. This automatically disqualifies this pump as we cannot crimp stainless steel tubing onto the discharge port.
Pump 3.png This was our second choice of pumps. The benefits of this pump is that the materials are confirmed to withstand the radioactive environment. Also, the shaft length of this pump is compatible with our desired tank design. The disadvantage of this pump is that at 1/5 gallons per minute, the pump produces 16 feet of head. We can overcome this obstacle by placing valves in the tubing system. If that is not a viable solution, we can purchase a smaller impeller from Graymills.

Design Selection[edit]

Filter Selection[edit]


Overall, we are very pleased with the design of our filter and we think that it did a good job of meeting all our specifications. One thing to note is that we did go back to having the bypass come back into the top of the filter. Although in this design it does not actually go through the filter cartridges it just runs along the side of them so that it can help direct the flow. The top of the filter is made from a solid stainless-steel block which overall made it heavy. The filter housing is made from a 5-inch diameter schedule 40 pipe, and we were pleased with how it worked in our system. To mount the top of the filter to our filter housing and we had to come up with a unique idea. We created a flange system that was welded to both the top and to the filter housing. We made two identical flanges and welded them into place so that the top of the filter would be able to mount to the filter housing in four different locations. Overall, this is a well-done design and it did a good job of achieving what we hoped when we were initially designing.

Tank Selection[edit]


There were two main criteria that we wanted to meet with the tank. First, we wanted the tank to be easy to clean. Second, we wanted something that would be easy to manufacture. We believed that our first initial concept fit those criteria the best. The design was then altered when we learned that the pump had a longer shaft than what we were expecting. Later we had determined that we wanted to reduce the amount of welding on the tank to minimize places for leaks to occur. This led to the creation of a flat pattern for out first initial concept. However, later we discovered that our design would require tooling to be made in order for it to be bent. Which is what led to the final design having the side sheared off, inset an eighth of an inch and then welded back on.

Pump Selection[edit]


Looking deeper at the LV series there are many options to choose from. The pump that correlates best for the required head and system architecture is the LVM21A pump. This pump runs at 1/8 hp, has a shaft length of 8 ⅝ inch, and requires 115 volts. In addition, the discharge port of this pump is ¾ inch National Pipe Thread. Provided that our tubing is only a quarter inch, a ninety degree elbow with a reducer threads directly into the pump discharge port to be compatible with our tubing. One of the bigger challenges of this pump is the ability to mount it. The final pump mount design is a two part mount that bolts to the bottom of the motor body with ¼-20 bolts. The two part design is essential as the diameter of the impeller is the same as the motor body making it impossible for a mount to simply slip on over the impeller. As one can see from the above picture, the two part mount comes together around the shaft and is then bolted. Due to the length of the shaft, the mount is designed so that the bottom of the motor body is elevated 1.75 inches off of the deck plate. This allows the inlet of the pump to have clearance off the bottom of the tank. This clearance is essential to avoid cavitation.

Integration Selection[edit]


This is the final integration design with all the placing of the components. This design is elegant, easy to manufacture, and completes all the design tasks we need to reach the client’s needs. This design places the filter before the pump, much like the other designs, but does not have the bypass going into the filter. The filer will run completely gravity fed into a slanted tank as opposed to the flat tank of the other designs. The pump is places on the other end of the table at the end of the tank and pumps out into minimal tubing which allows for lower head loss. There is a T-joint at the maximum height constraint that allows for two values to be places on the ends of the T-joint. These values are for controlling the flow onto the grinder and creating a bypass tube for excess flow from the pump. This design is what we ended up building for our project.

The Final Product of Our Project




Although the design met all client specifications, there were some improvements that need to be made to our design before it could be used at INL. These changes were suggested to us by INL and are as follows: incorporation of flexible tubing, moving the bypass to enter the filter housing, changing of the type of Swagelok quick disconnects, varying the heights of the guide pins, and t-bar handles on screws. The flexible tubing was suggested to us during our last visit with INL. In the final concept, the tubing cannot be removed easily with the mechanical manipulators, which makes it almost impossible to remove sections of our system in the hot cell. Flexible mesh tubing is available on the market and can be purchased for around the same price of the current tubing, which is about $200. Using this tubing throughout the whole system is recommended, as the manipulators can move the flexible tubing out of the way if it becomes a hindrance to the working environment.

Team Information[edit]

Biography Discipline
Drew Fagan

Drew Fagan is a senior in mechanical engineering at the University of Idaho. His interests include vibration and acoustic analysis. He hopes to continue his knowledge in these areas with the Naval Surface Warfare center to try to help the submarine fleet stay undetected in dangerous waters.

Mechanical Engineering.
Jerry Kahn

Jerry Kahn is a senior in mechanical engineering at the University of Idaho. His interests are in design and nuclear power. Currently, his goals are to work in industry for a few years then return to school for nuclear engineering.

Mechanical Engineering
Thomas Moore

Thomas Moore is a senior in Mechanical Engineering at the University of Idaho. He is interested in machining and design, and hopes to work in a fast pace atmosphere after school. His current goal is to work in the field for a few years before returning to school to work towards being a professor.

Mechanical Engineering
Troy Sanders

Troy Sanders is a senior in Mechanical Engineering at the University of Idaho. His interests include design work and piping and HVAC systems. He is passionate about roller coasters and hopes to one day design roller coasters or work on hydraulic launch systems for roller coasters.

Mechanical Engineering

Document Archive[edit]

EXPO Resources[edit]

Team Documentation[edit]

Drawing Package[edit]