A conversation with Elon Musk

On 2019-10-14 3:08 PM, JF Mezei wrote:
On 2019-10-14 12:30, David Spain wrote:

That is really a remarkable comment. Not as bad? How do you know? You
building Starships?

I was noting that Musk compared the cost of raw material per pound of
material. It is a nice PR stunt but doesn't provide correct comparison
since one does not use the same amount of carbon fibre (by weight) as
you would steel to build a structiure of same strength.

Note: I didn't state that carbon would be cheaper, just that his
cpomparison exagerated the cost difference.


"With carbon composite, you need to cut the fabric, impregnate it with
high-strength resin, which can be difficult and then make 60 to 120
layers!

Modern structures lay not carbon fibre "fabric" but rather
pre-impregnated carbon fibre strands. A robot arm strings the strand
along in precise location/direction. This allows very procise strength
is specific location of the "object" being built where the machine will
striong more strands over a specic location and in various directions
(instead of the fabric that ahs just 90° strand crossing each other)

This sounds like a hand wave to me. Any idea how expensive that is for a
large object? Is that even feasible for a 9m diameter rocket? How does
that compare in cost to laying fabric? Numbers?

After the laying is done, the object is then cured in an oven. The resin
used for the pre-impregnation requires heat instead of a catalyst
chemical to trigger the curing process.

Some structures such as large wing surfaces still make use of fabric.
For instance, the Bombardeir X-Series (Now Airbus A220) uses fabric for
wing siurface, but teh IArbus A350 uses a pen "drawing" system to
populate the surface with fibre strands. The later takes much mroe
programming than hand laying fabric, but results is more spophisticated
and mroe precise construction.

Okay so for large surfaces what Tim Dodd says holds I presume. You did
note the 60X figure (cost of composite vs stainless to mfg.) in the quote?

That's probably why SpaceX's composite casting rig went from this:

http://compositesmanufacturingmagazi...rocket-bfr.jpg

to this:


https://www.teslarati.com/wp-content...c-1024x346.jpg




"There is a further reason for using stainless steel rather than carbon
composite – fracture toughness. Carbon composites are typically very
brittle, and small flaw can initiate cracks which rapidly spread and
destroy the structure.


Which is why the PROCESS is extremely important when making any
composite object since you want a process that will not have a risk of
imperfections, airbubbles etc. Hence my comment about this being great
when you need to produce a whole bunch (since work to program the robots
to do the perfect job is spread amongts many produced units whereas if
you are producing only a handful, then it is a lot of work to setup the
robots for only a few units produced.


Not just mfg. process but handling too. *Again* I re-quote from previous
post:

" ...A subsequent lengthy investigation determined that the case of the booster had likely been slightly damaged during ground handling, providing a flaw which initiated a crack which rapidly spread."

Noted: "...*slightly* damaged during ground handling" & during launch
not while on the ground. A structural failure in a composite due to a
defect caused by a mishandle while on the launch pad that lead to a
quick demise. If structural integrity of Starship depended upon hull
integrity of a composite rocket it would still fail, liquid fueled or
not. Possibly in a fashion and for reasons quite hard to detect while on
the ground. In this case it's not performance (weight) uber alles. I'll
take that trade of some resilience for weight. Again I'm not second
guessing Elon here.

I don't really think you are either. But I don't consider your point
that Elon is being too dismissive as valid. He had already spent big
money on tooling for composite mfg. before a big change in direction and
the switch to stainless steel. I suspect he had given this much, much
thought.

Dave