#6838 X-SHARC | Build Blog 2020&2021

Hello Everyone!

We are team 6838 from Istanbul and we will be sharing about our design, build and fix process in this thread. We decided to build a brand new bot for Infinite Recharge which will help our rookie students learn about the design and manufacturing process more in-depth. We will hopefully be posting a general update once a week and updates worth posting in between those intervals.

So here it goes…

Let’s get the basic questions out of the way first.

Info About X-SHARC

We are currently 20 students located in Istanbul, Turkey. We are a team founded in the second half of 2017. We attended our first (off-season) event and won the event.

In 2018 we competed in Istanbul Regional and got the Rookie All-Star Award.

In 2019 we attended 2 regionals, Istanbul and Bosphorus, won the Bosphorus event and two other awards. Later in the season, we attended two off-seasons and won one of them and became the finalist in the other one.

In 2020, we competed in Istanbul and Bosphorus regional again. We were ranked 13th in Istanbul and were eliminated in quarterfinals. In Bosphorus, we were ranked 10th and became the 6th alliance’s captain. We were eliminated by Alliance 3 whose captain was 7285, a friend-team of ours, who won the event. (Congrats!)

Why are we building a new bot?

We decided to start over from the very beginning. We chose to re-strategize, re-design, and build a new robot from the ground up. There are several reasons for this decision:

  • Help our rookie students have the chance to be more involved in a competition-ready robot’s design and manufacturing process and learn more in-depth compared to the build season when veteran members usually try to rush things.

  • Build a bot that we could potentially use the major parts in the replay of Infinite Recharge.

  • We do not have any technical mentors and we lack machining capabilities. Therefore, we want to explore the machining possibilities around us, perhaps through our sponsors or commercial options.

What is this thread for?

We know that open alliance threads are extremely useful for inexperienced teams or any team looking for inspiration, motivation, and knowledge. We want to share what we know and learn with everyone as we build!

As a team in Turkey, we face many problems including metric unit differences with the imperial unit using products, the delayed and rare shipments of parts from the U.S., and etc. We do not know of any build blogs by any Turkish teams. So, we wanted to share the way we handle this process and share our experience with every team around the globe and especially the international teams who might face similar issues.

32 Likes

7 & 8 June | Strategy and Subsystems Meeting

We had a 3+ hours meeting where we revised our strategy in Infinite Recharge and how we want to approach 2021.

Here are the key points from the meeting:

  • We want to only score from the protected zones unless we are forced otherwise.

  • We want to be able to shoot from the trench run and shoot straight up against the small triangle near the power port. (Similar to 148’s strategy with 118)

  • We want to have very quick cycles. Our actual 2020 robot was not very quick in ejecting balls. We had to feed them one by one. We want to change this and have a much quicker eject time.

  • We had many problems with our intake this year. We want to have an intake that won’t pull the power cells under the robot. Also, we don’t want to miss pcs through the sides of the robot.

  • We want to have a minimum of 5 ball auto and an explore 8-10 ball autos.

  • We want to empty our storage as quickly as possible when the endgame buzzer rings and go do a level climb. (An extreme goal here could be achieving a 10-second climb through driver-practice and fixes.)

  • We are not planning to do a color wheel mechanism until it is past-kickoff. We want to make sure our climb and shooter are near perfect.

We also had a subsystem voting and results were:

  • Trying to implement WCD (we have been a mecanum team for 4 years :smiley: )

  • 3 Roller Intake (like 118, 1690, new 6328 Intake, etc.)

  • Spindexer

  • Two Position Hooded Shooter (Pneumatics actuated)

  • Turret (We are not sure about having a turret yet! We had a very heavy turret this yea, so we voted for it. Results were %53 Yes and %47 No. So, this decision is not finalized yet.)

  • For climber, we want to try a telescopic arm. We are not finalized on if it will be fixed or actuated to be perpendicular to the ground.

  • We are also looking into how we can feed balls into the shooter from the spindexer.

15 Likes

Calendar Layout

We should first layout our plan for the next week, two weeks, month, and three months.
We have never been very successful at sticking to deadlines. We want to make sure we fit this calendar.

Our overall calendar basically looks like this:

  • June | Strategy and CAD practices, Learning about Best Proven Practices based on Strategy

  • July | Actual CAD Month, CAD-Review-Fix-Repeat

  • August | Sorting out the manufacturing process (including searching for manufacturing sponsors), BOM, Tidying up the workshop (if we can visit the shop by then)

  • September | Actually manufacturing our parts, Getting the required robot parts from the U.S., (fingers crossed for this one) powder coating the parts, new member applications, and interviews

  • October | Assembling everything and finishing mechanics & electronics by the end of October, and getting the basic programming done by the end of the month. (Includes basic driver controls)

  • November |Breaking & Fixing the robot until the 15th of November. Simultaneously adding more autonomous controls to the robot, like vision and rpm/distance calculation. Finish auto routines by the end of the month. Have a competition-ready robot by the end of the month.

  • December | Perfecting our auto routines and mechanical flaws. Getting a ton of driver practice. Have two different drive teams trained at least 15+ hours each.

How does the rest of June look like for 6838?
We are currently having 2 hour CAD Practice meetings every Tuesday and Thursday.
We also have 1-hour long team meetings every Monday.
We are having more meetings when needed based on team members’ schedules.

June TO-DO

  • Discussing our strategy as a whole team. (Done)

  • Have a final strategy decision with team leaders based on the voting results, financial resources, and doability.

  • Have our (newly-forming) CAD subteam have the skillset and knowledge to CAD a full-robot cooperatively.

  • Accept new members to the team

  • Meet with teams that inspire us and learn from them as much as we can.

  • Teach our new members about the basics of FIRST & FRC and also about specific areas based on their subteam.

  • Make a detailed research about the mechanisms we chose to design and build. Learn the best practices of teams who have done it successfully.

  • Create an online Design Handbook that our CAD members can use throughout the CAD process and share it with the whole FRC community!

Online Design Handbook?

So, we want to have a website that we can use to learn about tips and tricks about the subsystems we are designing. But, we want it to have actually a lot more.
So, this is what the online handbook will include:

  1. Subsystem Types’ Definitions & Information
  2. Design Tips and Tricks and things to watch out for
  3. Useful Chief Delphi Links
  4. Some video examples of subsytems.
  5. [Hopefully] Part Tracking System Embedded into the Website
  6. A tab spesifically about our team’s manufacturing abilities and information for designers about how the parts they design will be manufactured.
  7. Some surprises may be added on top of these if we can pull all of this off.
  8. Maybe link this website to Sharc-Scout which we released last year.

Here are some photos of the first prototype of the website:

You can visit the live website by clicking here.

Disclaimer: At the time of posting this no tabs are actually working only the visuals and animations are being done. However, the link will be continuously updated.

21 Likes

Meeting With Citrus Circuits!

On Friday, we had the chance to meet one of the best teams all around the world, the Citrus Circuits #1678! We had so much fun and learned a lot from you guys. Thank you!

We have discussed several topics and exchanged opinions about many things including leadership, outreach, design, programming, and etc. We learned a lot about how 1678 operates, what are some of their best practices and we also tried to share some of ours! Currently, we are looking for ways to improve our leadership and subteam dynamics with 1678’s insights!

CAD Program Update

As we have stated in the previous posts, we are meeting up 4 hours a week to practice CAD. As our CAD members have begun using a CAD software around 7-8 weeks ago, we want to make sure every member feels comfortable designing. We have done 2D & 3D technical Drawings, Assemblies, shooters, a WCD and an intake to get involved with the FRC ecosystem.

On the 11th of June, we had a 2.5-hour meeting instead of a regular 2 hour one. We discussed how we felt about the program and gave feedback about the program. One of the most common feedbacks were as listed:

  • Some say that they are not aware of some tools in OnShape. This causes them to lose time. However, knowing about tools could save them a ton of time.

  • Most people felt comfortable with the “modelling” side of things but said that they would love to learn more about the technical design side of things as well. Some of these things could be knowing about tolerances, securing shafts, knowing how designs will actually be manufactured, etc.

So, based on these we decided on having a total of 8 hours per week for CAD meetings. 4 hours will be dedicated to CAD practice as usual. The content will be revised to meet the requests. Also, we will have another 4 hours split into 2 hour sessions where we will have workshops about how mechanical design works. We want to have people from different teams give presentations which we also would love to publish through our YouTube channel!

Subsystem Decisions

As we are quarantined, we lack the opportunity to make prototypes between the subsystems we could not decide. We are looking into basically two major decisions:

  • Turret / Fixed

  • Spindexer / V - Funnel / A varation of snail-like Storage

On Sunday, we had the fourth and last meeting to decide these major decisions. Afterwards, we will begin searching for teams who have built the subsystems we decided. We will be looking into these major points in this research:

  • Useful Chief Delphi Links

  • Manufacturability and Price Point

  • Teams’ best practices about that subsystem

  • The things we can improve on those teams’ designs

Our recently formed CAD subteam is currently looking into these about the subsystems we have decided on. Our deadline for this is 26th of June. At the same time, we are trying to arrange meet-ups with some of our favorite teams and learn more about them in all sorts of areas including design choices.

Meeting Decisions

Final Subsystems

We had a final meeting in which we decided about the subsystems.
So here are the updated and final subsystems:

  • Implementing a metric WCD
  • 3 Roller Intake (like 118, 1690, new 6328 Intake, etc.)
  • V-Funnel (like 6328, 3847)
  • Two Position Hooded Shooter (Pneumatics actuated)
  • Turret
  • A telescopic arm for climbing
10 Likes

Have Turkey teams returned to do team meetings in your workshops?
Anyway good luck in the re-build process.

2 Likes

Management Update

Trello
The lead mentor from 7285, Sneaky Snakes @vatanaksoytezer helped us learn about a great management system named Trello. Being inspired by them, we set up our Trello boards for all the sub-teams. Thanks Sneaky Snakes and @vatanaksoytezer!

We have struggled with a couple of people trying to memorize all the deadlines and remind people about it for a long time now. We have tried different solutions like having master excel/sheets or etc. However, having most things on a single page dedicated to a sub-team really seems to be more efficient. We hope we can get used to using this system and maximize our team members’ efficiency and motivation!

CAD Sub-Team Updates

We switched to a new CAD practice schedule as we have explained above. On Monday, we had our first go with the Workshops. We did not have someone doing a workshop. Instead, we opened a Watch2gether room and watched the “Mechanical Design” workshop held by 1678 in 2016.
After viewing the hour-long workshop we discussed what are some of the best-proven practices we could take away from this workshop. Here are the results:

  • Don’t design Perfect, design Close Enough

  • Design & Communicate (One is not more important than the other!)

  • Looking into how we can use GrabCAD, or if we need to use it as we are using OnShape

  • Implementing a Part Tracking System (embedded into the website linked in posts above or having it as a Google Sheet)

  • (Even though this does not have to do with the CAD sub-team especially,) We decided to have workshops like these hosted by our students and anyone from any subteam will be able to attend.

Also, on the actual CADing side of things, we have done a WCD and a couple of other subsystems before. For the upcoming two weeks until July, we decided the CAD a very simple robot to get a grasp of the whole CADing process. The plan looks like this:

  • CADing a Feeder on Tuesday (Done!)

  • CADing a Fixed Hooded Shooter on Thursday

  • CADing a V-Funnel Indexer/Storage next Tuesday

  • CADing an Intake next Thursday

We did the Feeder CAD Practice with the following specifications:

  1. 1.5" compression on the power cell
  2. Redline geared up to a 1/12 Sports Gearbox
  3. Motor shaft’s pulley: 18t 15mm HTD 5mm Pulley
  4. Driven shafts’ pulley: 24t 15mm HTD 5mm Pulley

Our CAD team is practicing using only these to figure out the rest by themselves. Our goal is not to have a perfect robot CAD that is ready to be manufactured. We rather want them to understand the concepts lie behind the design process.

5 Likes

We haven’t returned to our shops yet. However, we are doing online meetings. We hope we can get back to our workshops in September if not earlier. If you see the overall plan above, we are doing everything before September virtually. Also, thank you for your kind words!

3 Likes

OnShape already has all the functions of GrabCAD built in, so there’s no need to try to use GrabCAD with OnShape.

6 Likes

It’s also important to consider that you don’t know any of the potential game changes yet - FIRST won’t be announcing those until kickoff and at that point you may need to make some significant modifications to your robot anyway.

TL;DR: Don’t go building another robot right now - it’s probably going to be a slight waste of time and money to do much of anything until FIRST announces the game changes in January.

3 Likes

17 - 27 June Recap

It has been a while since we updated the blog. Our members had some projects to do and most of us were busy throughout the last week. However, we also had a very productive week overall. We are still getting used to using Trello, but we started to see many benefits of using it. We highly recommend it!

We were planning to interview our applicants last week and have an online orientation/welcoming day this Saturday. However, we had to reschedule all of this for next week. We hope to send out the results to the applicants on the 29th of June and have the Welcoming Day on Wednesday.

From last year, we learned that we should be able to build relationships before we attempt building robots. We want to make sure everyone feels comfortable stating their ideas and giving feedback to others. Also, we learned that essential information about FRC should be given to new members and everyone should have a basic overview of the robot and how it works. We will be doing this later on with a series of workshops.

Online Welcoming Day

We will have a Zoom/Meet meeting where we will have various presentations for the new members. The overall event is planned to be around 3 hours. The presentations schedule will be like this:

  • “Welcome to X-SHARC!” by Ilgın (CAD Lead & former Captain) & Irmak Su (PR Co-Captain)

  • An Icebreaker Event (Maybe icebreaker.video could be a good option for this.)

  • “How Do We Operate?” by Burak (Team Captain)

  • Subteam Presentations by Subteam Members (10 Minutes Each | Programming, CAD, PR, Mechanics, Electronics)

X-SHARC | Game Night

Two members from our team created a Discord server merely for gaming. It includes browser games as well so the whole team can play stuff together. They have also set up a bot to be able to play Werewolf directly on the Discord server. We are planning to use this as an ice-breaker for the new members and build up closer relationships between our members.

X-SHARC | Summer Camp

We are planning to have a Discord server dedicated to teaching our new members. Here we will have workshops by our subteam members and everyone will be able to join the workshops they want. We might also record and share these if we can.

Currently, our subteam leads are preparing a curriculum. It will have links to many resources and an overall plan of how the workshops will take place.

Subsystem Research

Our CAD members were responsible for researching different subsystems. They have done a great job at shooters, storages, and feeders. They are still trying to wrap things up with intakes, drivetrains, and climbers. If they can, we will be adding every resource we found to the website mentioned above.

July | CAD Month

In July we are planning to have week-long sprints broken into two sections. We will have these for 4 weeks in a row.

Every sprint will have a CAD review by a veteran member from our/other teams. At the end of the week, we will have reviews from an engineer & veteran members if we can. This will also be the meeting where we will decide what we did good&bad during the week and what we should do differently in the next sprint.

However, this plan will be finalized this Sunday and we are still looking for ways to improve efficiency & learning.

5 Likes

Hi,
Sorry for the late reply. We are not merely building this bot for competitive reasons. More information about our reasons could be seen in the first post.

Also, as you have probably heard about the FIRST’s current announcements, we are really optimistic about the game changes not affecting as much as some would consider.

But again, changes could be done and still we see both situations as an opportunity to learn. We think this will be a great opportunity especially for our new members to practice design and leadership. Anyways, thank you for your input!

2 Likes

First has already said that unmodified 2020 bots can be used in 2021 with no issues. This statement alone tells you a lot about what kind of changes they are going to make (and thus the changes they won’t make). I see nothing wrong with starting the rebuild process now, it’s not like shooting power cells is going away, or the shield generator.

1 Like

CAD Month Update

We separated our CAD students into two main groups. The first group consists of four people and their job is to CAD the entire robot with its subsystems. The other group’s job is to CAD 3D printable parts and design a field replica that would be easy to set-up & disassemble and affordable.

1st Sprint Mid-Week Review

1. Climber

Things to change:

  • The thickness of gussets will be changed from 2mm to 4mm
  • Different stages will have different sized bearings to make it more compact
  • Constant Force Springs will be researched to see if it is possible to get them in Turkey.
  • Gusset design will be revisited to make sure they fit more compactly
  • A stopping mechanism will be added to the end of each stage

2. Intake

Things to change:

  • The plate geometry will be changed to something less round to help us prototype quickly if needed.
  • We will CAD as if all shafts are hex for the first sprint but for the upcoming weeks we will try to change to round shafts.
  • We will try to learn what the compression should be based on other teams’ experiences.
  • We will look into how we can mount pulleys and polycarbonate tubes to a round shaft.
  • Some weak points will be fixed with the new side plate design.
  • There will be two plates per side and we will hide the pulleys in between those two side plates.
  • Compliant wheels will be the on the shaft touching the balls and other rollers will be polycarbonate tubes.
  • The ideal speed will be calculated to make sure we can pick up balls while driving.

3. Drivetrain

Things to change:

  • A custom gearbox will be designed (it will be the same as micro toughbox)
  • Will look into if we can run the chains inside the tubes
  • Gussets do not seem very strong, will look into that
  • A bellypan might be necessary to hold the frame together more rigidly.
  • We will look into if we should have another tubing run parallel to the wheel mount tubes to make things easier to mount. (Else, it might be too hard because of the chains running outside the tubes)

4. Shooter & Turret

Turret design is on the wait until we get the shooter done.

Things to change:

  • Will search for the applicable angle brackets to use to mount the shooter.
  • We will look into how we can make our shooter compact and two-positioned while it is on a turret. We might use a window motor with mechanical stops or a piston to actuate it.
  • How many pieces there will be in the backing of the shooter and what should be their thickness?
  • We will research how many gears there should be to create enough inertia.
  • We will look into the ratio of the pulleys.

5. Storage & Feeder

We will look into the feeder once we get the storage and chassis done completely.

Things to change:

  • The slope can be higher to make sure balls are passed to the feeder faster
  • Think about how we can mount the inclined surface to the chassis. (Possible idea: have another tube on top of the tube near the intake.)
  • We can add 2" and 2.25" compliant wheels to the beginning and the end of the storage. We can try this easily as we already 2 of each of those wheels.
  • We will add a plate to hold the shafts in place.
  • We will research other teams’ best practices to make sure balls don’t get stuck.
  • We will see if we can get wider HTD 5mm belts in Turkey to minimize pricepoint. This can also help us with unjamming the balls.
  • The pulleys in the drawing will be replaced with HTD 5mm pulleys.
  • Hex collars or e-clips will be added to the CAD.
9 Likes

You can look at 2363’s drivetrain series
It will be easier to draw wcd if you use weldments to draw drive rails. 3506 has a tutorial about using weldments for wcd chassis. You can also add custom weldment libraries to your cad program.

1 Like

Meeting With Makers Assemble #5951!

Last week, we got the chance to meet with Makers Assemble from Israel! It was really fun to learn how both teams function as international teams. We discussed our outreach programs and the technical design process. Thank you guys!

1st Sprint Update

1. Drivetrain


Things Done:

  • Gussets are redesigned to look better.
  • Custom Gearbox is designed.
  • Added bumpers to the CAD

Things to change:

  • Bumper height will be changed to 130mm from 127mm.
  • Will add a belly pan.

2. DT Gearbox
This is a custom gearbox using the same ratio as toughbox micro 1:12.75.
We added metric bearings and a sprocket to drive other wheels as well. We can share more info and even the CAD itself if anyone is interested.

Things to change:

  • Will add an encoder mount.
  • Will look into if we should have a Talon SRX mount nearby the toughbox, due to the short data cable.
  • Might have to make another version of it to accommodate more room for our funnel subsystem.

3. Shooter
We had a basic shooter geometry CADed. However, we decided to go back and revisit the shooting angles.

Firstly, we decided that 37 and 80 degrees were fantastic choices. Therefore we created a sketch that can be seen below. We noticed that we can not have a flat bottom with this design to make sure it is two positioned. So, we had to add what we call a “neck” to the shooter. We also planned to add a small feeder type of wheel to help balls get through the “neck” of the shooter.

However, this design was a bit bulky and added more cons than its pros. So, we sat down to find another set of angles.

Below, you see our second sketch with 37 and 70 degrees. We did some basic calculations and found that 70 degrees would allow us to shoot from the edge of the triangle in front of the power port.

We watched some of 148 & 118 matches and noticed how 148 was able to shoot from near the trench zone somewhat more than they did near the triangle. We figured even with such a fast scoring machine like 118, 148 still could go near the trench and score from there. So, we thought we could think of shooting from the triangle as a backup plan that we won’t use that often. When we have an alliance with a fast&efficient scoring robot from the trench zone, we can shoot from the triangle. We also figure this would be easy to accomplish with a bit of driver practice.

Also, this design allowed us to have a flat shooter bottom, resolving the feeder and other bulky stuff involved with the “neck”. Here is the sketch:


Actuation:
We plan to use a piston outside the side plate to actuate the hood up and down. Placing them outside could be a good way to prevent the piston from getting in way of the ball or sticking out in the back.

4. Intake

Things to change:

  • We will replace the polycarbonate tubes with another one with a diameter closer to our compliant wheels.
  • We will design another version with round shafts for the polycarbonate tubes because we can not access 1/2" hex shafts easily.
  • We will replace the Sport gearbox CAD with the actual 1:4 Sports Gearbox and Redline.
  • We will see if we can mount it to the chassis easily.
  • Will change the compression based on the feedback we receive from other teams.
  • We might add a rod/shaft to space out two sides of the intake evenly to make it more rigid.
  • Will add spacers between the pullets and bearings to allow the bearing to turn freely.
  1. Climber
    For the climber, we decided that we should upgrade the gusset thickness to 4mm from 2mm as we were planning to use aluminum sheets. However, that resulted in having the gussets not fitting in and we had to increase our max box extrusion and the size difference between each of them. So, we decided to change the aluminum to steel and go with 2mm stainless steel sheets. This was a decision between compactness and weight. This will allow us to have more a more compact climber but it might be a bit heavier.

Things to change:

  • Spacers will be added.
  • Gussets will be designed back to 2mm.
  • Mechanical stops will be added to the end of each stage.
  • Other hook designs will be looked at and we will try to make it look more aesthetically pleasing.
  • The bearings will be changed.
  • Constant Force Springs will be researched to see if it is possible to get them in Turkey.
6 Likes

Meeting with Rolling Thunder #1511!

Yesterday, we had the privilege to meet with #1511! We learned a lot from them, and most importantly we had a great time! They gave us really good advice on how we could improve both ourselves and our community. We hope we can meet again with them soon. Thank you #1511!

4 Likes

What you’ve got here is called a “double-loaded” chain. Usually in a WCD the center wheel is directly driven and the corners have chains. This means that at most, a chain is only carrying half a robot’s worth of torque. If you out the gearbox on one end, however, there’s a chance that a single chain will take the entire robot’s movement torque.
It’s safer to run 2 individual strands of chain. One from gearbox to center wheel, and one from gearbox to other corner. We did this in 2019 and liked it.

You should do an FEA on the climber hooks. I think aluminum would be fine.

Looking like a solid robot, excited to see what’s in store.

2 Likes

Thank you for all the kind & encouraging words! Also, thanks for the technical assistance, as a student-led team, we really appreciate technical insights!

As on the double-loaded chain, that is a really valuable point we think we can improve on. Our main problem with running chains overall is the fact that it makes running tubes perpendicular to the tube with wheels super hard. This also makes the overall mounting of subsystems a lot harder. We think of running two more tubes parallel to them or adding tubes on top of them. Adding them to the top would make it a bit harder as we are building a low-bot, but it seems like the best option.

We are a bit afraid if we make the gearbox thicker, we won’t be able to run parallel tubes and be forced to have tubes on top of chassis tubes. We really think running them as two separate chains could be really beneficial but we could really use other ideas about the problem I explained earlier. Are there any easier ways to mount subsystems? I think learning what you guys did in 2019 could help us.

1 Like

Storage Update

Currently, we are using two redlines, one for each side, connected to 1:4 Sport Gearboxes. On the output shaft we have another 1:3 reduction with 15 and 45 tooth pulleys which will be 3D printed, resulting in a 1:12 reduction overall. We had a great experience with 3D printing them this year but we never 3D printed such small or large ones.

The two supporting shafts on both sides are not 1/2" hex shafts, they are 15mm 3D printed shafts which we will run threaded rods through and nylock nuts on both ends.

We also plan to add two more supporting shafts parallel to the holding plate extending from the plate to make it even more rigid. We might reduce the overall number of them.

We also want to explore if we can run the motors on the opposite side to make more room for the telescoping climb. (This will probably conflict with our DT-Gearbox.) We might also just take the motors further away and still run the pulley to the end of the funnel with a larger belt. We might keep both versions and see which one fits the overall design more.

4 Likes

I’m not sure what you mean. Two chains will only be very slightly thicker and shouldn’t affect the overall design. They would replace the 2 chains you have now.

This doesn’t mean doubling up one chain. You run one short chain from corner to middle and one long chain from corner to corner.