Seans Mechatronics Design Blog @flamingoutcast-blog - Tumblr Blog

Reflections on a project...

This project has really pushed me to think outside of the box and also think in a way to better the group. While large portions of this project have been exciting and I am ultimately happy with the final design; it has been a somewhat tedious project that felt at points as if it slowed completely to a halt.

I figure It is best to start with the negatives and move to a more positive tone towards the end.

I feel as though for parts of this project organising meetings for 5 people became a bit of a nightmare as we were limited to only before or after our classes during the week that we could really have meetings. While we still made it work; having 3-4 hours of Mech Design in a row is quite taxing resulting in loss of focus. Speaking on this some of our meetings slowed completely to a halt just due to conflicts in ideas not particularly being resolved. This most heavily effected the concept selection side of things as it became very difficult to find anything that we all agreed on.

However, in saying this and for all of our trouble in making decisions; when we made decisions they were great. While indecision can lead to time wasted and delays, for us they actually all turned out benefiting our final design and in some ways I am quite glad to have actually spent more time on things(bar the stress). We also occasionally hit spots where we were all on the same page and it was just brilliant, this was most prevalent when we were creating our morph charts as we just sat their for hours ideating and feeding off each other until we had it done. It was fun, it was productive and well worth the time spent.

I feel as though the big things I have learnt from this are to make sure that everyone is on the same page about what our plans are. The second is to not wear the black hat too often, instead to try questioning than just trying to tear peoples ideas down.

I am glad to have the group that we did even if things were hard at times and Everyone all contributed equally to this project :)

Validation and Verification

The last thing our project required was to look at ensuring our designs right from the start of our project could be traced into our embodiment design and ensuring that we could meet our user and system requirements comfortably.

Monique put together this document below and it is very detailed and self explanatory so I will leave that to document attached and also as this is my last post enjoy a hedgehog climbing into a sock!

https://docs.google.com/document/d/1iqa7QGLvjejNjyWeXc3PuVWD6Pd3Ug3pyzefL60IjSI/edit?usp=sharing

Fun Extra Cool stuff!!

So another thing I really wanted to be able to do was to see our model in action and so I quickly made up a very rough model of how the whole thing would fit together. It is missing two wheels, the slide cant actually slide and is at very best rough but I thought it would be a good visual representation of where we have come to :)

Embodiment Design!!!!

It is quite exciting to finally be at the point where everything has led up to the embodiment design of our product. I compiled all of the individual parts that everyone had done and was in charge of doing the sketches of our final designs. I realised however that we had not completely thought of a way to ensure that the servos would stay in place or particularly how they would actually be held.

Taking a bit of creative liberty(The team was OK with this) I decided to actually fix the servos to several points moving up the shaft and use corkscrew gears(shown below) to move the segments up and down. This was just to ensure that we had lots of teeth contact between the servos and the gearing, reducing stress and keeping the movement fluid.

The actual servo and container have been connected in the drawing below. so that the servo can stay entirely stationary throughout the entire process this allows for a full 180 degree extension of the upper arm.

Below are the rest of the sketches of our embodied design; annotations have been made on the pages and I will continue after the pictures.

It also became clear that the initial mechanical pushing of the slide by the person using the machine would not sufficiently fit the brief. It resulted in too much stress and pushing down metal beams is not something that happens naturally.

Instead we adapted to two more servos(shown below) that run along the bottom of the track to move the sliding panel in a controlled manner and allowing them to return to resting position easily.

cleverorchaoticdesign

My final concept contribution

Developing the Head.

The beads method needed to be tested so we came up with a prototype to demonstrate whether it would work or not.

Click the image in the post above this one to see it working

Proof! the beads have a good chance of working!

However I realized there was a potential mode of failure.

This picture demonstrates that in the second step and as seen in the previous posted video, that at the time when the sock is going around the heel on the bottom of the foot it travels a larger distance than the part of the sock that is on top of the foot. As the beads are all traveling at the same speed and covering the same distance this would mean the sock would not go all the way around the heel and up the ankle before the top part of the sock is fully rolled out.

I found a solution to counter balance this design flaw. The beads travel half the distance than the piece that pushes them because the beads have to fold in half back on themselves. If a couple rod’s ( that do not double back on themselves) are fitted on the bottom we could push the sock up and around the heel quicker whilst only rolling the top part of the sock a little way up the foot. These rods would be made from a strong and flexible “Shape-Memory-Alloy” that would remember its shape enabling it to curve up and around the heel.

I also thought that if the beads were held in place by railings either side of the beads then the apex of the beads would actually protrude above these railings. The beads would then also drag the sock along ever so slightly when they are being pushed forward mitigating some of the frictional forces created by the constricting sock on the outside sleeve.

my finished concept solution for the head piece which holds, inserts and guides the sock over the foot too.

The shaded area at the back of the sleeve represents a pressure pad sensor that detects when the users foot is in place inside the device and when it is removed. This helps add to the non-functional requirements as it is safely activated and deactivated by the users foot and can measure the pressure applied to the foot. Making it activated by the foot also increases autonomy and increases the speed of the operation, the foot activates the device automatically when it is inserted into the head.

Here is my final concept design of the head.

I added little rollers between the rods for comfort reasons so they would roll smoothly under the users foot and around the heal.

flamingoutcast-blog

I thought It was important to include this as Nick explains the concepts behind the head really well and also the potential failure modes of the supporting rods that would bend up around the ankle.

Nick mentioned that we would be using a tensile “shape-memory-alloy” specifically NiTi which is a relatively ductile material that can be easily compressed. There is not a lot of information regarding its compression other than it has a significantly lower fatigue failure rate. However this is still 20000 cycles which in operation would be more than sufficient. This still doesn't necessarily account for constant compression however and as such the material may be less reliable in that sense. However developments into “shape-memory” alloys are likely to be the best solution to this in future.

Sensing a pattern here, too much Pressure

Just to briefly talk about sensors I wanted to touch on where we would have sensors involved in our design.

1. The most important place for us to have a sensor is in the head of our device just around the support ring. This is important so tat the machine can know when the sock has been successfully applied to the headpiece. This will likely be 3 or 4 simple LDR’s that signal to the Microcontroller when they have been broken.

2. The Infrared sensor in the Roomba base; the aforementioned sensor in the Roomba base would likely need to be a medium range High sensitivity sensor in order to aide movement.

It is also important that I talk about the pressure plate inside the sleeve on the head of the design so that the machine will be able to recognise when a foot has been inserted and to accordingly lower the servos to a comfortable level where the foot lies with minimal force(this will more that likely reach the base before this point).

Control, and automation

This will just be a brief post but I thought it would be good to talk about the different steps of operation of our final design.

This mainly outlines all of the steps the machine will have to undertake along with the points at which the user will have to have input. We decided a simple remote with On/off, Start, and height buttons would be best to control this process. For any stage in bold to progress to the next(except on) start must be pressed.

General process:

Initial input Move off changing pad (pad one) Locate base pad (pad two) Stabilise Wait for input (ready state) Secondary input (lift) Machine lifts up User puts sock on machine Wait for input (height- up to 500mm) Third input Machine moves down and holds position User lifts foot Machine analysis of angles Machine adjusts User inserts foot Machine slides beads Machine feels resistance (20N) Machine moves away from the foot (minimum clearance of 50mm) Machine waits for input (new sock or OFF) Another foot: Repeats process from second input Finished: Moves down to ready state Disconnects of base pad 2 Moves to changing pad (pad one)

Hand remote buttons are: START LOW MID HIGH (3 height options) OFF

This was intentionally extensive to show how much the robot would be doing alongside how little input would be required from the user. If any stage is left idle for 5 minutes or longer the machine will turn off and return to the charging pad.

Roomba in the house! Ambitious Steps

One Important thing that I have forgotten to touch on as of yet is the Roomba base. We are highly pulling from the Roomba design(cross section visible below) and just adapting the wheel exchange so there are four retractable wheels for stability.

The Roomba Design is very intricate and the vacuum sections in yellow would be replaced with the mechanisms for the rotating arm. Other than these changes however the design would stay pretty much the same this is due to the ability of the Roomba to find its own way around the room with an infrared reciever.

Similarly we talked about the design being able to find its own way between a charging Pad and an operational pad one where it can wirelessly charge and the other set-up in an ideal position so that the user can comfortably use it. While this is a cool concept however we are unsure of how we will be able to ensure it follows the same path every time and so instead are unsure if any technology currently would be sufficient to perform this task. However what we ideally imagine are an infrared receiver on the design that allow it to find and easily locate the operations pad and then when operations are finished; the charging pad. This sensor would effectively have to be of medium range and high sensitivity in order to accurately find the location of these pads.

The one other thing I felt important to bring up was how we are planning on stopping the design from moving. This is highly based around the concept of a shower mat; either with small suction cups or even just a rubber material it would be able to lower down onto the operational pad similarly made of a rubber material and this would create high friction and stop lateral movement. This is ideal as when pressure is applied to rubber surfaces they become harder to separate keeping everything stable.

Software Desicions

In terms of our Software choices we didn’t particularly have the time to really look at anything specific but instead decided to look at the steps that would be involved in the programming to get an idea of how it may operate.

Nick wrote up a list of how these operations may play out in a system

“

Operational steps:

-The user is seated

- The User calls SockBot to users position with the remote

- SockBot Roomba function triangulates and moves to users position.

- Once stationary at users position Roomba stops and raises its wheels

-Sockboat raises the head ready to receive a sock

- LDR detects when the user places the sock over the head

- Once LDR detects sock is over the head it lowers back down to the base

-User places foot inside the head and activates pressure switch

-The sock is rolled and pushed onto the users foot

-User removes foot releasing pressure switch

-SockBot raises it’s head to receive 2nd sock, the process is repeated until the second sock is on the users foot

- After application of second sock and the users foot is removed SockBot retreats away from the users position.”

These while very simple would be the basic pseudo code as to how our machine would operate and the different states that it would be in. It is likely the code would be written as a state machine with many different Switch statements controlling operation. The Roomba base will likely be controlled via the same Microcontroller also leading to having a central Microcontroller for the majority of operations.

FMEA for our Design

Now we have had a look at our final design and before we moved into the final stages of embodiment design we wanted to do an Failure Mode and Effects Analysis in order to get an idea of how our Machine would operate under stress. It was incredibly important as we needed to make sure that for all of our moving parts we had an idea of how we could incorporate contingency Methods for our design.

Mitch Spearheaded this and designed the table.

This was an important step for us as it allowed us to look at whether software or Mechanical solutions would be more ideal for each of these failure modes. It also brought to light that due to the design and high stress placed on the servos and base; there was potentially going to be a lot of overheating that we would have to consider.

Complexity and the DSM

Our Design at this point was beginning to really amp up in complexity and as such we decided that A Design Structure Matrix would be very valuable for us to assess what is entirely dependent and how we would have to be careful with minor adjustments coming up to our embodiment design.

After Assessing the DSM it became clear that we had large segments of our design that were entirely independent front one another. For example if we were to look at the Head and application parts they were completely unaffected by changes in the base or the Lifting Mechanism.

However 3 problem areas arose.

1. If we were to change any of the servos it would affect the power supply and in turn effect all of the other servos. Ensuring we have a suitable servo selection is detrimental.

2. Any Changes to the base or Roomba style design would in turn effect the wheels, the movement and also the stability of the system so it was important we got this right.

3. Any Major changes we made to the head piece and applicator would heavily effect the ability for the sleeve to slide and the design of the foot rest.

Practical Design

The Next step allowed us to have a look at some practical design. We Split solutions up into several sections. Nick looking at the actual applications side of the sock; specifically the head and sliding bead Mechanism. I was to look for Mechanical Solutions and Mitch was to have a look at electrical solutions. Sven was also to look at the Design of the external shape during this process. My potential solutions for the Mechanical solutions can be seen below.

These specifically look at several ways we can approach both the turning of the base and also the lifting Mechanism of the machine with its embedded components. As I designed these I realised how impractical it was to simply explore just the mechanical side of things as we would have to incorporate some kind of Servo or electric component anyways but it was good to get an idea of how it would actually have to operate. Explanation for each of the ideas are similarly on the pages next to them.

(More Text below these pictures)

As I continued designing these I realised how impractical having small hydraulic supports would be and that if we could get a system that operated without them that would be ideal; alternatively just having them be completely solid.

Mitch looked Quite heavily into Servos and implementing them into the design of the arm using an AT90USB1287 MicroController. The PWM function of this would allow for controlled movement in all of the servos. This MicroController has 4 seperate PWM lines and as such would be able to support the 4 Servos required at different points in the arm, base and head.

This was decided as it would be more valid, easier to program and incorporate and ultimately would fit in comfortably as a Mechatronics Solution.

Secondary Prototyping

One of the things we discovered with Sven’s Design is that there ultimately was going to be extremely high stress within the folding Mechanism of the sock. The sock is initially rolled over the entrance and behind a flap. Once the foot has been inserted into the entrance and pressure is felt on the foot rest the flap would fold up, over and around the ankle to apply the sock. While this is a nice design in theory it provides too much stress on the sock either damaging the sock or potentially creating a dangerous situation for the user.

Instead Nick came up with the solution of potentially using the mechanical qualities of beads on a wire to move the sock while still keeping it close to the foot.

We mocked up a small prototype to show the cross section of this operation:

This would replace the folding panels and allow very little movement of the sock laterally instead keeping the movement up the foot.

Concept Selection

For our next step after prototyping we decided that we should compare our individual concepts to those of everyone else in the team. As we each had 2-3 concepts we figured that it would be fitting to limit them down to the most practical 2-3 before doing a “Harris profile” analysis to determine our final concept.

The group of us met up and shared our designs(meeting minutes ahave already been posted in a previous post) and along with my designs here are some of the rest of the teams concepts:

Monique:

Sven:

Nick:

It was quite exciting to actually get everyone's designs together and do an analysis of how we thought everything would fit together. After some careful deliberation we ultimately limited it down to Nicks design, Svens folding mechanism and Moniques eggbeater idea. This was mainly due to discussion over the practicality of the stress required to keep the sock open.

Instead of using Pugh’s model to evaluate further we decided to look instead at a model called the Harris profile that Sven had used prior. The model uses a series of pluses and minuses to evaluate whether something is excellent, ok, sub-par, or terrible.

After doing a final evaluation Sven’s design came out on top, and we decided to look at further developing on this idea. One thing with Svens model however was that he had only designed the application method and not the base. This led us to spend a bit of time evaluating our options for the base and how it would best suit the applicator.

For the base we decided that stability was key and that we would look at developing a solution that would be as stable as possible. My design(below) was decided on as the triangular prism base that down sloped to the bottom allowed for us to keep the stability in the event that the machine was to tip. Keeping the base wider than the operational area also allowed for increased stability again as all the force will be centralised.

We also looked at using the cherry picker rising mechanism from Nick’s design as it gives us the most adjust-ability and control over different heights and angles.

A few Changes...

While looking at the concept of Changes I found that it was important to address a few of the big changes that happened during the course of our concept design and prototyping.

Firstly we decided to cut donning from our brief, while it would be nice to have a machine that both dons and doffs socks there is too much time pressure to design a multi-purpose machine that can easily do both, and it is in best interest and to get the best design possible to cut this.

Secondly we changed the brief to eliminate the standing portion to again simply our product. While it would be nice to be usable while standing we feel it will be safer and more efficient to have operation seated and will less stress on the machine.

The User requirements that this takes out of play are:

- The user must be able to doff sock.

and the System requirements are:

- The system must be able to grasp a sock that is already on a foot.

- The system must guide a sock during removal for the socks full length.

- The system must move clear of the foot once the sock has been removed.

- The system should allow easy access to the sock once it is removed.

and the test criterion this affects:

- While the syswtem is in operation it must be able to withstand -700 Newtons statically loaded while the user is standing.

This was voted on and is in the best interest of the team to provide an efficient product within the time constraints.

Initial Prototyping

One of the things that we discussed while exploring potential solutions for “SockAid” was the actual Practicality of a mechanical system being able to emulate the motion required. This motion involves holding the sock open; wide enough for a foot to fit in, and then to move up and around the ankle to release the sock. There are a lot of steps involved in this and so we had a look at how we take our socks off and also how we could emulate this with a simple system.

The three videos seen above all represent out different efforts to see how a sock is doffed. Mainly looking at the important zones we can see that it is possible to take the sock off from the base without any pulling but impractical to try and pull it without any support from the back. Sven showing an efficient way to take them off, myself trying foolishly to pull it off with one hand and Nick explaining some of the different processes

This led to our next set of tests which involved exploring different ways that we could emulate this motion. We ultimately came up with 3 prototype sections we wanted to try. 2 for donning and 1 for doffing.

The first of these, is looking at the idea of suspenders and having 3 leads pulling the sock open in all directions. We simulated this with actual suspenders. However we found that the resistive force of the sock increased exponentially when it was being pulled on from every direction and as such could not keep it open for long enough to be a viable option.

For our second donning test the second we looked at using a mechanism to hold the sock open with a semi solid design that would have small wiggle but ultimately stay solid. We emulated this with a section of a toilet plunger stand; pre-placing the sock on so Monique’s foot may be inserted. It worked well and was a relatively crude design but as a whole it allowed room for the foot to fit while keeping it a lot more stable and allowing a good application. Unfortunately I didn't get a video of the actual application but below you can see the setup at the very least.

Our final prototype explored the single finger design that Nick talked about in his video above. Using the single finger approach to move down the back of the sock and pull it down past the ankle we were able to successfully remove the sock. We emulated this movement with the toilet brush taped to a plastic hook.

Concept Solutions

While I have already posted the minutes for this concept I just wanted to take a step backwards and look at my own concept designs and then to explore a little bit about the rest of the teams designs

The next logical step moving forward from the morph chart was to look at how our idea could be developed into a complete solution. We all decided to go away and come up with our own concept solutions that would fit in both with our requirements specification and also the morph chart we had created. For my two designs I decided to focus on developing designs that were complete and had all components thought out, taking inspiration directly from the morph chart.

My First design looks at having a machine, that is very easily movable with a ‘rumba’ style moving base and retractable wheels to make sure the machine is entirely stable during operation. This would be paired well with a static or pressure mat to increase the friction with the floor. A back brace runs out from the center of the base that is connected to a horseshoe up the top of the brace. The sock can be easily fit inside the horseshoe and the edges rolled over in order to hold the sock open.

The foot rests on a freely moving raised spring pad that; as the foot enters the sock and moves downwards, also moves downwards and is streamlined into a hole that only fits the pad in a particular direction. This ensures that the foot will always be in the right place during operation. When the pad has reached the base and the sock has been pulled off of the horseshoe (in correlation with the test criterion stating if 10N of force are applied the sock it must release); sensors signal the horseshoe to split open and allow your foot to be removed.

My second concept looks at the importance of a more stable product. with a triangular prism at the base to ensure that most of the weight is focused towards the ground. Multi-directional wheels line the base in order to allow for ease of movement to any position on the floor. two hydraulic support beams compress down as force is applied to the sock and the horseshoe until they have been fully compressed and the users foot is on the pad at the base. The Pad then senses this and begins to slowly lower a trap door style platform towards the ground allowing the foot to slide comfortably off. The machine then slides away.

In the next post I will explore the other team members solutions and the challenges that we faced in designing and ideating coherent solutions. Along with a brief look at some prototyping.

Trending Blogs

Recently Viewed Blogs

Seans Mechatronics Design Blog