We need to orbit the satellite back.
So our next talk is a continuation on the theme of Libre Space.
Now we are going to talk about the the the deployer.
So please welcome Thanos.
Hello everyone.
So I'm Thanos.
I'm also a core contributor at Libre Space.
I'm a mechanical engineer too.
And I will continue the talk of Vilias basically and that's some stuff about the
satellite's final safe house, the deployer and specifically our own deployer, the pick-up
bus.
So some words about the deployer.
What's a deployer?
So now that our satellites are like the cubic small, you need to have a way to attach it
to a rocket.
Okay, so basically as mentioned before, we build a box, place all the satellites inside,
mount the deployer on top of a rocket.
Here you can see multiple deployers actually.
And when the time comes, the rocket gives a signal to the deployer and says, okay,
open your doors and deploy the satellite.
That's what the deployer does.
So how do you start designing a deployer actually?
So that's a really tricky question.
For us, we knew that we wanted to house 8P units.
So 8 of the cubics, for example, that you can see.
So I will do a quick walkthrough of the internals of the pick-up bus just to have a bit of context
and then we will dive deeper inside the pick-up bus itself.
So you start with a rail.
You place all the satellites on the rail.
Some way we will see how.
Then you need to push the satellites outside with some kind of spring most of the times.
And then you need all these things to be mounted on the rocket.
So you put the flange there to be mounted on the rocket.
After that, it's a best practice to close everything because space and also add some
kind of door to keep the satellites inside.
So now you have a box with satellites inside that can be pushed outside.
But you also need a locking mechanism and deployment mechanism actually.
So when the rocket gives the signal, the satellites do not stay inside.
They go in orbit.
So that's the final version of the pick-up bus, V1 actually.
And these are the main components of it.
Let's dive deeper now.
So again, we start with a rail.
That's one of the most basic parts, but one of the parts that actually restrict you because
of the satellites.
The interface of the satellites and the deployer too.
So we have the pocket cube standard.
This is the pocket cube standard actually that gives you the dimensions of a pocket cube
and it's available everywhere around the pocket cube too.
So satellite manufacturers can put side panels there, sometimes even deployable things, which
is cool, which is the hot stereo you can see around.
So for the pocket cube, there's a base plate down where the deployer actually interfaces
with this base plate.
So we manufactured the rail actually.
So the rail, that's the top view of the rail.
So you see here it has like a notch where the satellites slide inside and are held from
the base plate.
So for our specific rail, it was machined out of space grade aluminum, 7075 T6 alloy.
It also was hard to analyze to give the surface as much strength as possible because satellites
were sliding inside this two millimeter slot.
So now, yeah, sorry.
Yeah of course.
So now we're moving to the PUSHER sub-assembly, which is what pushes the satellites outside.
So this is a really early version actually.
We actually try to follow the rapid prototyping procedure as much as possible.
So we constantly 3D print parts, break parts, then redesign parts, and then 3D print again
parts and then break them.
So after much discussion, we opted to use constant force springs.
This is really good.
It's a good practice because you cannot just take a proper compression spring and just
compress it all the way.
So it gives a really big range actually.
So with a really quick paper towel calculations, we got a rough estimate of the spring strength
and also the satellite exit velocity, which is a really important number when you're
building a deployer.
So when we finished with the paper towel calculations, we actually machined by hand
a dummy rail, as you can see there.
We 3D printed PUSHERS and barrels and attached the spring to do some testing.
So let's see the PUSHERS of assembly in action.
That's our first prototype.
And it seems to be working.
Yeah, and they did the drop of the table.
That was a really good one.
So it worked.
So we moved forward with this design actually.
So again, you can see here the pick up us.
You can see one side of the rail.
So we have the same assembly mirrored.
So we can house 4P on the one side and 4P on the other side.
You can see the PUSHERS of assembly here.
And you can actually see the machined part in the middle, the final rail with the PUSHERS
that's now machined from PTFE material.
It's not 3D printed.
So the PUSHERS was made out of a single billet block of Teflon.
That Teflon is a really great material because it's space great, which means it doesn't out
gas.
It can operate really good in volume.
It also has a really, really small friction coefficient.
So it slides amazingly with a hard-denodized aluminum.
The second part was a barrel.
It can be seen in the right side of the photo, which the course and force spring was wrapped
around the barrel.
And the spring was attached on the top of the rail.
So in the right picture, you can see the PUSHERS that are in the top.
That's the deployed state of the PECO bus deployer, actually.
So moving on, we now go to the doors of assembly and the thermal knives mechanisms.
So, okay, now we have the rail.
We have the PUSHERS of assembly.
We need to design the door, too.
So you need two things.
You need the mechanism to hold the door close.
And you need the mechanism to open the door.
That's the mechanical thing.
You also need, from the electronic side, some way to attach the door and then cut it with
a signal from the deployer.
So we used this mechanism.
These are early prototypes.
So we used the pin puller mechanism with a compression string.
And in order to hold the pin in place and actually secure the door closed, we used the
enema string.
And in order to cut the enema string so the door would open and the satellite would be
deployed, we used thermal knives.
Basically, a nigh-chrom wire which gets heated, cuts the enema string.
The spring is decompressed.
The pin is pulled.
The door opens.
So we did more prototyping.
And we knew that we had to build electronics from scratch, of course.
It was mandatory.
So the PCB that we did would handle the communications with the rocket in order to receive the signal.
We would also have the thermal knives attached to it and also the deployment switch.
Of course, you need a way to see if the door actually opened.
So yeah, we used two thermal knives because space and redundancy.
And we also used two enema strings.
So only one of the thermal knives needed to work for the deployment to happen.
So you have two.
It's a more redundant system.
So that's the final door subassembly with the machine parts and everything.
So you can see here the enema string is wrapped around like that.
And the two thermal knives are here, one here and one here.
So the two strings hold the pin inside.
They get cut.
The doors get released.
So the door subassembly was complete, finally.
It was ready to be integrated to the rest of the deployer.
So behold, that's the deployer, the final deployer, the final pick-up bus.
You can see here the wire harness.
This is getting connected to the rocket.
So when the signal is received, the door will open.
So now let's see pick-up bus in action, actually.
So this is the deployment test that we did.
And you will see the thermal knives starting to glow.
There they are.
The pin will be pulled back.
And then the door will open.
And satellites will fly outside.
In this specific case, these are dummy-mass satellites.
They're not actual satellites.
They're blocks of aluminum.
But yeah, that's how it works, actually.
And you can see the push-up assembly that now it reached the top there.
So continuing now, I want to have some slides about the testing that we do and how do you
space grade an assembly, actually.
So one of the testing things that we do is called protoflight testing.
Protoflight comes from two words, prototype and flight hardware.
So when building pick-up bus, we had a really short period of time to do everything and
build the two satellites that would be inside.
So you had six months.
So protoflight helps you with the time that you have developed really easily.
So in this case, the qualification model, so the model that is tested and goes through
the vibration and stuff, is the same as the flight hardware, actually.
And the satellites were integrated inside pick-up bus when this protoflight campaign happened.
So basically, you want to see if the DQT, the device under testing, will sustain the
launch.
Actually, launch is a really, really bad time for the deployer and the satellites.
Huge vibration, huge accelerations, everyone.
You really need to see and be sure that balls will not start flying around, pretty much.
So the steps are the following.
So step one, you do a resonance survey.
What that means is you identify the eigenmodes of the system.
Actually, you need to have, usually, the first resonance should be pretty high, at about
100 to 150 hertz.
But that depends on the launcher, actually.
So you get these tables.
You place the whole deployer into a machine, a vibration table.
You vibrate it on all axes, and you get the resonance frequencies for each axis.
Then you do a signed vibration profile.
So this is where the bad things start to happen for the deployer.
So you basically start punching the deployer with vibrations and the satellites inside
and hope it survives and doesn't break.
So you pass from five hertz to 125 hertz with a signed wave profile.
That's really painful to watch.
But what's even more painful is the random vibration, which is step three.
That's the real heating.
And you pass at the same time, the same machine, from 20 hertz up to 2 kilohertz.
So that's the profile, and the deployer must sustain this.
When random vibration finish and you are a bit relieved, OK, things seem to go OK, you
took what's static testing.
We simulate the static loads exerted on the deployed unique launch.
Which again is painful to watch.
But when everything is finished, you do again another resonance survey.
So you add another table, basically, to there.
So you have the post-serve resonance.
So you need these values, the resonance before the testing and the resonance after, to be
the same.
If they're not the same, something happened.
Something flew through, something broke.
In our case, they were the same, actually.
Some words now about our design simulation tools.
So everything we do is open source.
The tools we used to do.
Our things are open source, too.
So we use free cut to do all the modeling, everything.
We also use free cut to do all the simulations.
Before starting to hit your deployer in the vibration table, you need to do simulations.
So we do a lot of model simulations to actually try and predict the eigen modes before sending
the pick up bus to be hit in the vibration table.
We also do static simulations because the deployer is bolted from the flange, and you
have a lot of stress, stress is going around.
We use calculus as a solver to do the vibration in free cut, and we also for the electronics,
we use key cut to do everything.
So as I mentioned before, the pick up bus was developed in a really short time.
So now, eventually, we had a bit of time to develop the version two of the pick up bus.
So by using the tools that we mentioned, we actually were able to do a lot of improvements
that were obvious with the pick up bus we won.
So the most, the improvement that mattered the most was the mass improvement that we
did.
So because it was designed in such a short notice, we had a lot of big safety factors,
and we tried to do it really quickly.
So now, we had more time to do simulations and more time to do design stuff.
So by iterating the plates, we actually managed to cut the mass in half of the deployer.
So right now, again, the capacity is the same.
It can house up to eight satellites inside.
It's almost half the mass of the pick up bus we won.
It has a larger satellite envelope, which means you can fit more stuff around the satellite,
but it has a smaller deployer envelope overall.
It's a bit smaller deployer.
And again, it has updated electronics because we found some minor issues, so we fixed them.
So we did a really cool thing, again, with the isogrid patterns that you can see there.
It's really space rated.
So we had an aluminum frame this time, not just a plate.
And we closed it with a huge PCB, an FR4 PCB, and polyamide tape, couple of tape, pretty
much.
So we closed it, and it was secured.
We will see how that goes.
We will start manufacturing the next months, and the expected loss of this is in the next
few years.
So before we leave, if you were watching about the pick up bus story, so pick up bus one
was integrated, as Ilya mentioned in the previous talk, and it was launched, actually, but the
launch, again, didn't go as planned.
Yeah, it was a really sad thing to watch, but space, these things happened, explosion
happened.
So, but there was a plot twist.
We received a phone call from the guys at Firefly, and they said, sometimes, we think
we found your payload.
We were like, what do you mean, it just blew up at 15 kilometers.
You cannot do this, it's waste.
What do you mean?
No, no, no, we think we found your payload.
We were like, yes, but so they found our payload.
So that's the pick up bus that survived through a rocket explosion at 15 kilometers.
And there's more.
It was okay.
We opened it, removed the satellites, they worked.
It was, everything was so okay that we thought they didn't launch it.
The only thing broken was this electronics cap and this poor capacitor.
It was amazing.
It was like, it didn't even work.
Like it was great.
And they actually sent us photos of the pick up bus mounted on a carbon fiber part from
the payload bay, and we think that this huge carbon fiber part reduced the drag, and they
found it in a beach.
Mear van der Beek's Spaceport Base, when it was launched, it was like in the middle
of the sand, two meters in front was sea, two meters from the back it was rocks.
We have photos, but we cannot share them unfortunately, because they don't let us, but it looked
pretty much like this.
So yeah, pick up bus survived because space is hard.
Pick up bus seemed to be harder though.
I will close with this really cool video, which is the orbital pick up bus.
This again is provided by Firefly.
They had the camera on board.
So let's just enjoy it for some seconds.
Open source space, everyone.
So unfortunately, because of the launcher they didn't have, as Ilias mentioned, they
didn't have the doubt, they didn't download the open pick up bus.
So that's what we get.
Which by Firefly is great, like pick up bus.
With the air.
So yeah, that's it for me.
Yeah, that was another deployer actually.
So real funny, really quickly, when we were watching the stream, and there was a cut in
the stream, and every day after that at the time said that they had confirmation that
all three payloads were deployed successfully, and the stream came back and we saw this,
and the door was closed.
And we were like, no, come on.
But yeah, after that, the Cubics worked and beaconed.
So yeah, successful deployment, everyone.
So I will keep that there.
And I'm open to questions if you want.
Thank you for the talk.
Why is so much solid structure on the outside even required?
So the question was why so much solid structure is required?
So it's not really.
In the pick up bus V2, you can see it's a more lightweight, but the pocket cubes are light.
They're 250 grams, but when you have eight of them, it's a bunch of kilos.
So when you place it in the rocket, acceleration, it's a really hard phenomenon.
So you need actual structures to hold the rail in place so it doesn't move in place.
It doesn't move everywhere.
So the plates on the sides of the deployer actually held the rail in place.
So you also need to sustain the vibrations too.
So you need to have a bit of a mass to absorb these vibrations too.
In the pick up bus V2 actually, instead of having a solid plane, we have two ribs.
I can show you in the presentation.
You can see it actually.
We have the design is two ribs that secure the rail pretty much.
So here is attached to the rail.
So when you have the whole rail protruding out, you need to support it like that.
So we did a lot of way saving by doing this.
And you also need to have it separated from outside from everything else.
So you need to have the deployer closed.
So in case a satellite has a malfunction or it breaks or whatever,
the deployer contains the satellite malfunction too.
So you need to have it closed.
Did I cover your question?
Yeah, thank you.
Thank you very much and congratulations.
Oh, welcome.
My question was in a eventually version three or version four,
which you consider also having a communication bus to communicate with the
satellite swatter inside the deployer for system checkouts, for example.
So we don't know yet, to be honest, it would be a case.
I don't know, we can ask the electronics guy.
Have you ever thought about this?
So yeah, maybe what it would be beneficial for
the other version was to provide the chance of charging the satellites.
So for the people that put the satellites inside, whether it's us or
other people to have like a way to charge them after the integration,
that would be an extra step too.
About communicating with the satellites, I don't think so because I don't know,
but I don't think so because when they are inside,
there are key switches that are pressed.
So the satellites do not start becoming or deploying or whatever.
In the cubic, you can see the two switches on the bottom plate.
So when they're inside, they completely shut down.
Yeah.
Yeah, here.
Yeah.
What kind of simulations did you perform?
Raise your hand, I cannot see you, sorry.
Yes, here.
What kind of simulations did you perform and
did those simulations inform further iterations of the design?
Yeah, all they did.
So much.
Like we do static simulations to simulate the static loads and
see if they can withstand these loads.
And we also do the model simulations that I mentioned before because in order to
do the vibration testing, we have to fly from Greece to Spain.
So we need to be sure that, sure.
We need to have a measure to see the eigenmos and stuff.
So we do vibration testing and static simulations.
So here it's a very good example because we did the through simulation,
we reduced the mass actually of the part.
Hi.
In your test deployment video, when the door opened,
there was a certain amount of bounce.
So did you have to take precautions to try and make sure that the satellites didn't
hit the door on the way out?
Yeah, so the springs, in the door there is a hinge mechanism,
so it has two torsional springs and are really, really strong.
So the back glass is minimal.
Now in the V2, we are thinking about adding a locking mechanism to it so
the door doesn't bounce back.
But it's not such a huge problem, to be honest, because even in the open position,
the pick-up bus door has a bit of preload because of the torsional springs
themselves, their legs are a bit angled.
So even in the open position, there are like half a Newton meter of torque.
So it stays there.
Welcome.
Yeah, there was a front question some here.
But yeah, yeah, okay.
Hi, thanks for the talk.
Did you know why the door closed after the deployment?
Oh, it didn't.
No, it didn't.
It was, it stayed open.
You mean when it's torbid, correct?
No, it stayed open.
It didn't close.
So why did it look like it was closed?
Yeah, because the Firefly Stream was cut and we didn't have a footage of the door
being opened.
We only have footage of the door being closed before the satellites came outside the deployer.
No, no, it stayed open, but we just don't have footage because the stream was cut from
the rocket, the stream.
They didn't download this specific part.
Yeah.
Do you, do you dump the vibrations in any way from the rocket through the pick-up
bus into the satellites?
As far as I know, no, we certainly don't.
I don't know if there is any such mechanism in the payload bay of the rocket itself.
But as far as I know, no, we certainly do not.
We have time for one more question.
Hi, thank you for the talk.
I just wanted to ask what science did you do?
Was this just proof of concept to show it would work?
You mean for the deployer, I assume not the satellite itself.
So, oh, sorry.
What science?
The deployer is pretty much a box.
So it has one work to do, deploy the satellites outside.
So there is the aspect of the proof of concept and the mechanical functionality of this thing.
We managed to build it and make it the RL9, so it worked in space.
So that was the goal, to have this kind of box work in space and deploy the satellites.
But the science behind it, it doesn't have the experiment, maybe.
You can say that the experiment was to deploy the satellites, actually, if you want.
Yeah.
Okay, thank you for very interesting talk.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.
Thank you.