CAD Workstation

[markten]

Nope - no hanguns whatsoever. Not even the Police carry guns. I think there is a UK made AR15 type gun but it is single shot and in fact you can buy genuine Kalashnikov rifles but again, in single shot.

hHere are only two places in the UK big enough to shoot .50 BMG, they are all military ranges which we share when they are not busy. The competitions are 1000 yard but the biggest problem then is half the year here you can't see a 1000 yards! Not like you crystal clear desert air. You can see the range by visiting this link:

http://www.fcsa.co.uk/index.htm

I'm using rifles as a means to learn CAD as they are quite simple, I'm not doing anything comercial but it does occur that an injection moulded all ceramic military rifle would be cheaper and longer lasting than a metal and plastic one. It would be like the equivalent of a Glock as it would never wear out and be equally indestructable. But there are too many gunds in the world!

 

[bob4432]

Originally posted by: markten
Nope - no hanguns whatsoever. Not even the Police carry guns. I think there is a UK made AR15 type gun but it is single shot and in fact you can buy genuine Kalashnikov rifles but again, in single shot.

hHere are only two places in the UK big enough to shoot .50 BMG, they are all military ranges which we share when they are not busy. The competitions are 1000 yard but the biggest problem then is half the year here you can't see a 1000 yards! Not like you crystal clear desert air. You can see the range by visiting this link:

http://www.fcsa.co.uk/index.htm

I'm using rifles as a means to learn CAD as they are quite simple, I'm not doing anything comercial but it does occur that an injection moulded all ceramic military rifle would be cheaper and longer lasting than a metal and plastic one. It would be like the equivalent of a Glock as it would never wear out and be equally indestructable. But there are too many gunds in the world!

that is a nice range. nice pics, i like how the "nemesis" has the buis on it for some close quarter stuff....would be pretty messy though here the ranges look like brown earth, not even black, healthy earth, just beige/brown earth. a lot of the pubic ranges are 200-600yds but not 1000.

to bad everything has to be single shot - is there a rule to only have 1rnd capacity too? for 50bmg, 338, 375, 416 or the larger rounds that is ok, but '06, 308, 223 or just about anything else you should be able to shoot more than 1 at a time, but i guess laws are laws. so is most crime over their people beating others up? or is murder done with a knife or rope? i know the us has probably the worst crime stats in the industrial world, but i don't think the uk is immune. just curious.

do the cops cary tazers or mace? i couldn't imagine a leo without a weapon.

i know glock and anybody else has that makes polymer weapons are making a ton of $$$ on their weapons, but they work and that is what matters.

 

[Howard]

You can't injection mold engineering ceramics AFAIK, but you can use powder processes.

Is it really worth it to make everything out of ceramic? I could see the barrel benefiting from the higher stiffness and less tooling operations, but the receiver and all that?

 

[markten]

Originally posted by: Howard
You can't injection mold engineering ceramics AFAIK, but you can use powder processes.

Is it really worth it to make everything out of ceramic? I could see the barrel benefiting from the higher stiffness and less tooling operations, but the receiver and all that?

Think it depends on when you start to define the nature of the precursors but in ceramic injection moulding a plastic mix of a ceramic powder is pressed through a constricting die to produce elongated shapes that have a constant cross-section. The powder mix consists of a fine ceramic powder with the appropriate additions of binders and plasticisers to give the desired flow properties, either cold or when heated prior to being forced through into the die. The organic binder slowly burnt out in a controlled atmosphere by means of a carefully controlled heating schedule, prior to sintering. Injection moulding ceramics can make series production very economical. You could conceivably injection mould almost a complete rifle in one go.

It's only a curiosity and a design exercise for me, steel rifles have been perfected for over a century and any improvements a ceramic can give would be only incremental. Having said that, target shooting is all about consistent accuracy and I think it's worth a try.

I still haven't figured how to make the tool for a barrel yet. I want to mould the rifling but then extracting the tool that does that seems difficult. I might leave the foundry to figure that one out...

I want to make all the rifle from ceramics as its a design exercise but it would also eliminate differences in specific heat and harmonics in using different materials i.e. a ceramic barrel on a steel receiver. It will also be much lighter, will never rust and can be cleaned in a dishwasher for a sparkly finish

I didn't explain too well about the single shot thing. As far as I know, pump action shotguns are limited to three shell but rifles can have as many as you want i.e. a full magazine. What you can't have is any kind of auto or semi automatic rifle so if you have an AR15 it must only be capable of firing a single shot, then you have to cock it again. Lee Enfields used to be made locally to me and it's said a British Soldier could fire accurately, over 60 rounds per minute which seems pretty fast to me. There's lot of them about in .308. My dad had one in national service but was never given any bullets!

We have all the crime you guys have and in all its forms. The cops have guns if and when they need them and pepper spray and tazers. There?s lots of knife crime, most kids carry knives and plenty of fighting but that's mostly drink related. The gangsters seem to have all the guns they want too, just last night a 19 year old was shot dead on the steps of the local prison as he was just leaving from visiting. Having said that, I live in Liverpool which is in some respects, the New York of this country and there has only ever been one instance of a cop having shot someone. And that was a mistake.

Generally we just abuse each other with our wit and cynicism.

 

[Howard]

Right, right, forgot all that I learned.

Couldn't you use a wax plug for the bore and rifling and compress the powder around that?

 

[markten]

Originally posted by: Howard
Right, right, forgot all that I learned.

Couldn't you use a wax plug for the bore and rifling and compress the powder around that?

that's what i'm hoping can be done, almost like a lost wax technique but I don't know if it's compatable with the other parts of the process yet which is why i might unload the problem onto the foundry. I find when talking to experts that's there's nothing they havn't encountered and solved before, much like these forums which are so helpful to people like me with little knowledge of computers.

 

[bob4432]

Originally posted by: markten

Originally posted by: Howard
Right, right, forgot all that I learned.

Couldn't you use a wax plug for the bore and rifling and compress the powder around that?

that's what i'm hoping can be done, almost like a lost wax technique but I don't know if it's compatable with the other parts of the process yet which is why i might unload the problem onto the foundry. I find when talking to experts that's there's nothing they havn't encountered and solved before, much like these forums which are so helpful to people like me with little knowledge of computers.

i would say the riflling would be the hardest part to figure out. i have no knowledge of what you guys are talking about but with metal barrels, the rifling is a "super secret" for match type barrels. how exactly would you figure out what profile to make the rifilng? or would you just pick up a nice, high quality sub moa barrel and copy its rifling? also, how would the ceramics deal with the distortion and or whip when a round is fired and 50,000psi(for a smaller round, not sure what 50bmg is) is generated in an instant, are the ceramics you are talking about able to handle that easily? this has been a very informative discussion

 

[markten]

hey Bob - you hit the nail on the head there with the rifling. I intend to do what all other genius designers do - copy form others! Lucky for me all this has been worked out before for each caliber and ballistic properties. Off the top of my head I think a .50 BMG is 8 grooves and about 1.25 twists per metre of barrel length but I'd have to look it up again.

As for "whip" or harmonics, that is one of the defining characteristics of a ceramic barrel. Ceramics are orders of magnitude stiffer than steels. If you treat it as a cantilever you can model these chateristics and they are substantial. Another reason I need a new computer!

I think 50,000 PSI is about right for a .50 BMG so I'm looking for a margin of twice or three times that. I'm looking for software that models pressure vessels.

Some other great features are very low coefficient of friction and chemical resistance which may mean higher velocities per load and little or no accretion of the metal jacket onto the interior of the barrel thus helping accuracy still further. Even better, the barrel could be cleaned by vapour/sand/grit blasting with a suitable tool in about two seconds flat! As long as the blasting medium is less hard than the ceramic it won't even mark it.

A ceramic rifle would be impervious to temperature change and other climatic changes too. The barrel won't change dimension this side of 1000oC.

Probably the best property of ceramic is in compression with failure at about 3500 MPa. The bolt would therefore be mega lightweight.

I'd also like to try a motorcycle wheel out of ceramic but I don't understand the loads suffciently and can't identify any software that would be an aid.

The easiest way to make a barrel would be by extrusion but this will have to be tried...

 

[bob4432]

very interesting. i have never really gotten into precision shooting, but do the actual groove make a difference? i understand the twist rate, but the actual groove and it dimensions and angles.

with the weapon being able to save weight everywhere you are going to need to design one hell of a muzzle brake or risk broken collar bones

as far as loads can't solidworks do it? i know it can do it for various types of metals, but i have only played around with it. one would think you could put in the properties of your materials and then allow solidworks to test it out.

also, ever think of doing sound supressors this way? once you do a little research into them, they are not that hard to understand and make work, like i said all i am lacking is a lathe and i would have at least one on a walther p22 for the price of the tax stamp.

 

[markten]

i'm not to owell up on the number of grooves either. I want to avoid a break. the ones i've seen shoot dust and other crap back in your eyes or your scope so I'm thinking of incorporating a telescopic stock with a spring damper to absorb the axial load. If I can balance the centre of mass in alignment with the barrel line it should kick straight back and not up & over.

One reason I want Inventor so much is that it too can give you information on material and loads. The problem with CRC's is the lack of data to do FEM so mostly it will be done by probability analasys.

If I can make the .50 BMG work I want to make a 6mm PPC and 6BR which are more practical and the most accurate calibers of all.

I've never really thought about suppressors or looked into them.

 

[bob4432]

Originally posted by: markten
i'm not to owell up on the number of grooves either. I want to avoid a break. the ones i've seen shoot dust and other crap back in your eyes or your scope so I'm thinking of incorporating a telescopic stock with a spring damper to absorb the axial load. If I can balance the centre of mass in alignment with the barrel line it should kick straight back and not up & over.

One reason I want Inventor so much is that it too can give you information on material and loads. The problem with CRC's is the lack of data to do FEM so mostly it will be done by probability analasys.

If I can make the .50 BMG work I want to make a 6mm PPC and 6BR which are more practical and the most accurate calibers of all.

I've never really thought about suppressors or looked into them.

good luck on doing it without a break, i have never seen a 50 without some type of break but would surely like to. what weight do expect it to be? 20#s or less? i am sure others would like to see it too and am sure if you can get this to work you may be some interested people, unless your company already does certain types of "contract work". the benefits of the ceramics sound nearly too good to be true, i wonder why nobody is making them already. hell, even say on the ar15 which over here in the u.s. has a huge user base a ceramic barrel with the specs you are talking about would be greatly appreciated. about the only thing nearly as "new age" is a very thin steel barrel wrapped in carbon fiber. a total ceramic one that has the low friction, doesn't suffer from heat would be a soldiers dream. sounds like you are putting in a great deal of science into this and this excellent, hell you may even change the theories of how modern small arms are built if you can keep the gun coming back straight so follow up shots in semi/full autos would be very easy and very accurate - again certain parties would be interested.

definately let us know how it goes

 

[Howard]

Originally posted by: markten
A ceramic rifle would be impervious to temperature change and other climatic changes too. The barrel won't change dimension this side of 1000oC.

CTE is still non-zero at low temperatures.

 

[Twinpeaksr]

SCSI is great if you have a number of drives on the BUS, this will ensure that it is faster than SATA or PATA, but also much more $$.

I would recommend RAID5 for Data storage, good fault protection, good speed. I have run both RAID 1 and RAID 0 before, I would not recommend RAID 1 due to the lack of high risk of lost data.

Right now I have been working on my setups trying to figure out what to do. I am looking at a 2 system solution and a single system solution ($2500-3000 range for either configuration). The original plan was a servier with ~1TB RAID 5 running Linux on a gigabit back bone, and then the work station with a single program drive. Dual core makes sense for me as I have many application open at one time.

I too am coming back up to speed, my last major system work was when the Athlon was king and there were only about 3 chipsets for the Athlon. That was also when Intel could not get processors made that worked, I hope that has changed.

Always good to get ideas of what others are doing out there, most are building for gaming, CAD and Engineering Analysis workstations are just not discussed enough!

 

[Howard]

Originally posted by: Twinpeaksr
SCSI is great if you have a number of drives on the BUS, this will ensure that it is faster than SATA or PATA, but also much more $$.

I would recommend RAID5 for Data storage, good fault protection, good speed. I have run both RAID 1 and RAID 0 before, I would not recommend RAID 1 due to the lack of high risk of lost data.

You wouldn't recommend it because there's a lack of high risk of lost data?

 

[Twinpeaksr]

Opps, how about high risk of data loss. The mind runs faster than I can type!

 

[bob4432]

Originally posted by: Twinpeaksr
Opps, how about high risk of data loss. The mind runs faster than I can type!

you are meaning raid0-striped, raid1 is mirrored so 100% redundancy...

 

[bob4432]

Originally posted by: Twinpeaksr
SCSI is great if you have a number of drives on the BUS, this will ensure that it is faster than SATA or PATA, but also much more $$.

I would recommend RAID5 for Data storage, good fault protection, good speed. I have run both RAID 1 and RAID 0 before, I would not recommend RAID 1 due to the lack of high risk of lost data.

Right now I have been working on my setups trying to figure out what to do. I am looking at a 2 system solution and a single system solution ($2500-3000 range for either configuration). The original plan was a servier with ~1TB RAID 5 running Linux on a gigabit back bone, and then the work station with a single program drive. Dual core makes sense for me as I have many application open at one time.

I too am coming back up to speed, my last major system work was when the Athlon was king and there were only about 3 chipsets for the Athlon. That was also when Intel could not get processors made that worked, I hope that has changed.

Always good to get ideas of what others are doing out there, most are building for gaming, CAD and Engineering Analysis workstations are just not discussed enough!

wouldn't a CAD/Engineering workstation just be a high end machine (gaming) with a different video card? i am assuming with those types of machines you need a lot of cpu cycles, good ram bandwidth and a good disc susbsystem.

most people here run a 32bit version of xp, usually home or maybe pro, not a lot run 64bit versions of windows or linux so must of the discussion is going to be limited to that.

 

[Twinpeaksr]

Boy am I off today, yes I was refering to stripping as being less redundant and more prone to problems.

A gaming system is a good starting point, but the focus will be more on the disk subsystem and CPU than on the graphics. There are also many different types of workstations for different tasks. If you are doing FEA or simulation, raw CPU and lots of RAM are the key, if you are focusing on CAD and 3D modeling, the disk subsystem and GPU are going to be more important.

The key lies in what you are trying to do, and with all the choices, I believe figuring out the type and speed of the CPU is the main decision. Single Core, Dual Core, Mulitple CPU, 64/32bit, this will be heavily dependent on the type of work you are intending on. It is easy to build a fast system, but is it the fastest for your application is the trick of it.

It was so much easier when all you had to do was pick a clock speed!

 

[wwswimming]

Adaptec 29160 with a 10K Seagate Gen 5

... i have this in a Xeon system, 2.4 GHz on a Tyan motherboard.

retired.

the 2 160 GB Parallel ATA drives that i have in a Shuttle system are
faster.

on my P4C800 system i have a SATA 120 and a SATA 300 GB. faster even
than the 2 160 GB drives.

i really like SCSI too. i have 3 SCSI systems. i guess i'm just saving
them for heating duty. or because i like the sound they make.

 

[markten]

with regard to mobo's I settled on this one for a workstation:

http://uk.asus.com/products4.aspx?l1=3&l2=82&l3=249&model=1289&modelmenu=1

...however, I read over the weekend that the app I want it for (Inventor) have plans to go 64bit, possibly in their next release so that means if I am to future proof I need to go back to sq 1 on find a two CPU workstation board which works with C2D/Xeon's. At this point I think it's wiser to wait a few months until there are more on the market. Or buy a MacPro

 

[bob4432]

Originally posted by: markten
with regard to mobo's I settled on this one for a workstation:

http://uk.asus.com/products4.aspx?l1=3&l2=82&l3=249&model=1289&modelmenu=1

...however, I read over the weekend that the app I want it for (Inventor) have plans to go 64bit, possibly in their next release so that means if I am to future proof I need to go back to sq 1 on find a two CPU workstation board which works with C2D/Xeon's. At this point I think it's wiser to wait a few months until there are more on the market. Or buy a MacPro

unless i am totally off, going 64bit doesn't mean they are going to mult-core support. having a quad setup would kick a$$, if the application used it, if not 1 core would be used very highly, a second would be doing the background and three and four would just be hanging out. again, check into this as i could be wrong.

 

[markten]

What about this new iMac? The spec/price at $2550 is pretty close to the system I specked using an E6600/Asus P5WDG2 WS-Pro and softmodded GPU. It's a little slower clock speed on account of using a mobile CPU (memron?) but it looks like it could be a decent work station and comparable to the the above spec. Of course you are not comparing apples with apples (sorry!).

 

[Twinpeaksr]

Biggest problem is it does not support the apps and I can't stand Mac OS. So far the system I have setted in on is about the same price, but includes 1.2TB including 960GB in a RAID 5 array and dual 20" LCD Monitors. The cost vs performance vs compatibility, the IMac just can't compete. Thanks though!

Proposed Workstation:

1 MSI K9N Platinum Socket AM2 NVIDIA nForce 570 Ultra MCP ATX AMD Motherboard
1 AMD Athlon 64 X2 4200+ Windsor 2.2GHz Socket AM2 Dual Core Processor
1 mushkin 2GB (2 x 1GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit
1 MSI NX7600GS-T2D256E Geforce 7600GS 256MB GDDR2 PCI Express x16 Video Card
1 HighPoint RocketRAID 2300 PCI Express SATA II Controller Card
5 Seagate Barracuda 7200.10 ST3320620AS (Perpendicular Recording Technology) 320GB 7200 RPM SATA 3.0Gb/s
1 NEC 16X DVD±R DVD Burner Black IDE/ATAPI Model ND-3550A
1 Lian Li PC-60BPLUSII Black Aluminum ATX Mid Tower Computer Case
1 Thermaltake TR2 W0070 ATX 430W Power Supply
1 Athena Power BP-SATA3141B 3 x 5.25"Bays to 4 x 3.5"HD (SATA) Back Plane
2 ViewSonic VG2021m Black 20" 8ms LCD Monitor
1 Arctic Silver 5 Thermal Compound
2 BELKIN AV20100-06 PureAV Digital Coaxial Cable, 6ft, Black
1 Microsoft Wireless Optical Desktop 1000 BV3-00003 Black USB RF
1 Tripp Lite Smart1000LCD 1000VA UPS

 

[bob4432]

Originally posted by: Twinpeaksr
Biggest problem is it does not support the apps and I can't stand Mac OS. So far the system I have setted in on is about the same price, but includes 1.2TB including 960GB in a RAID 5 array and dual 20" LCD Monitors. The cost vs performance vs compatibility, the IMac just can't compete. Thanks though!

Proposed Workstation:

1 MSI K9N Platinum Socket AM2 NVIDIA nForce 570 Ultra MCP ATX AMD Motherboard
1 AMD Athlon 64 X2 4200+ Windsor 2.2GHz Socket AM2 Dual Core Processor
1 mushkin 2GB (2 x 1GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit
1 MSI NX7600GS-T2D256E Geforce 7600GS 256MB GDDR2 PCI Express x16 Video Card
1 HighPoint RocketRAID 2300 PCI Express SATA II Controller Card
5 Seagate Barracuda 7200.10 ST3320620AS (Perpendicular Recording Technology) 320GB 7200 RPM SATA 3.0Gb/s
1 NEC 16X DVD±R DVD Burner Black IDE/ATAPI Model ND-3550A
1 Lian Li PC-60BPLUSII Black Aluminum ATX Mid Tower Computer Case
1 Thermaltake TR2 W0070 ATX 430W Power Supply
1 Athena Power BP-SATA3141B 3 x 5.25"Bays to 4 x 3.5"HD (SATA) Back Plane
2 ViewSonic VG2021m Black 20" 8ms LCD Monitor
1 Arctic Silver 5 Thermal Compound
2 BELKIN AV20100-06 PureAV Digital Coaxial Cable, 6ft, Black
1 Microsoft Wireless Optical Desktop 1000 BV3-00003 Black USB RF
1 Tripp Lite Smart1000LCD 1000VA UPS

why the switch to amd? the conroes beat the x2s in just about everything, even single threaded apps...

people will probably comment on the power supply and suggest a fsp(fortron), enermax, pc&p, seasnic or enhance in the 450-500W category.

 

[mleonard]

I just skimed this thread and saw very little mention to the video card! This is a 2D/3D CAD workstation and the GPU is very important! Almost as important as the processor. I noted a few refernces to the Quadro FX 3500 which is a great High-End card, but pricey at approx $900-$1000. Some prefer you to overclock GeForce gaming card, bad idea for a stable workstation. Professional cards are built to run with high-end applications like most Autodesk/Discreet products including AutoCad Inventor, and they have professional certifications.

You should seriouly consider the nVidia Quadro FX line or ATI Fire GL cards. Each have Mid-Range cards for approx $500-$600. With a single one of these cards, there's no need for SLI.