1-d pipe/tube flow simulation
J. Todd Wasson
I'm trying to write a simple computer simulation of airflow in a straight pipe.
Currently, I'm dividing the pipe into a bunch of cells, then keeping track of
mass, pressure, and density in each cell (temperature is assumed to be constant
for now.)
Then, I'm defining the boundaries between the cells seperately, so far, these
just have one state that is currently used: the mass flow rate.
In advance, forgive me for being uneducated in this area please :0)
In the simulation, I calculate the pressure (based solely on density for now)
in each cell. Then, I run through each boundary, calculating the pressure
difference between the two cells that share each boundary. This is multiplied
by the cross sectional area to find a force across the boundary.
Once this is done, I multiply the force by the timestep to get a momentum
change (impulse) across the boundary. This alters the boundary's mass flow
rate directly. Each cell's mass is then altered according to the total mass
flow rate from both of it's boundaries...
I'm sure this is overly simplistic, but I was hoping that the resulting
pressure waves would propagate at the speed of sound. Occasionally they do
this, but only if the cell sizes are set "just so". Right now, the pressure
waves will reflect off the ends as rarefraction/compression waves
appropriately, and interact with each other realistically, but the speed of the
waves is usually far from correct, and seems more dependent on the cell length
than anything else. Does anyone have a simple addition/suggestion to get me
thinking in the right direction here?
One concern that I have is that I'm not using the momentum within each cell for
anything, but rather I'm trying to alter the mass flow rates at the boundaries
directly instead. Is this wrong? I've been searching for information online,
but have not found anything explaining this yet.
Thanks in advance!
voy...@ihpc.net
"J. Todd Wasson" <jtw6...@aol.com> wrote in message
news:20020220185031...@mb-fw.aol.com...
What is 1-d pipe? I think pipe has 2-D geometry. What means "pipe" for you?
Alex
J. Todd Wasson
dimension, i.e., towards one end of the pipe or the other. No motion towards
walls of pipe, no swirling, turbulence, etc.. Flow across each boundary is
assumed to be constant, rather than having higher flow rates at the center of
the pipe..
J. Todd Wasson
wave usually does not travel at sonic speed as it should:
http://performancesimulations.com/files/pipe0001.exe
(43kb, 32bit Windows only)
To end the program, close the window with "x" at top right.
Thanks,
Todd Wasson
---
Performance Simulations
Drag Racing and Top Speed Prediction
Software
http://PerformanceSimulations.Com
My little car sim screenshots:
http://performancesimulations.com/scnshot4.htm
voy...@ihpc.net
orthogonal of speed of you media, that create parabolic profile of media
velocity. I see 2D problem in your description. Does it correct?
Alex
J. Todd Wasson
>orthogonal of speed of you media, that create parabolic profile of media
>velocity. I see 2D problem in your description. Does it correct?
>Alex
Perhaps that is the best way to do it, but I'm looking for a simpler
approximation for now. I've heard Navier-Stokes is very difficult to use in
code, and I have no idea where to begin with that :0) I have no calculus
background at all.
What I am concerned with primarily are momentum effects from pressure waves
bouncing between the ends of the pipe in this application. I am thinking it
will be fine to assume the velocity is constant at each cross section/boundary
seperating two cells (for my purposes), so I would like to neglect such a
parabolic speed profile across the boundary.
I described this as a 1-dimensional system (I could be wrong though, of
course) because I'm only allowing motion in one direction. There is an
x-velocity (towards or away from the pipe ends), but no y/z velocity in the
cells. Each of these cells is merely a cross sectional slice of the pipe.
In the past, I made a simple approximation that applied a force to the mass
currently in a cell, which accelerated it. Given the new velocity, I'd
calculate how much of this mass would move to the next cell. Given enough
cells, this worked pretty well for what I was attempting to do, although an
engineer would probably balk at the idea. Regardless, the pressure "waves"
moved at the speed of sound like they should, so it worked well for what I was
doing.
Now, however, my "bright idea" was to use the forces at the cell boundaries to
change the momentum, or mass flow rate, at the boundaries themselves, instead
of using the individual cell masses. This is where things are not working
properly, I think.... ?
Thanks very much for your previous response, do you have any other ideas, or
am I misunderstanding what you're referring to with the Navier-Stokes idea?
Todd Wasson
voy...@ihpc.net
pipe, where Navier-Stokes does!) by pressure gradient, correct? That is very
well known problem, if I correctly remember my 4th student year: describe
gas flow from area of high pressure with boundary conditions: a) free
surface from both sides of high pressure area; b) wall from one side and
free surface from another one; c) there are 2 areas of high pressure and low
pressure area between them; etc. All these cases, I guess, could be solved
analytically by characteristics. Probably, you faster do solution by
yourself then look for solutions in student's textbooks. You don't need to
develop the code, you described, just grub pen and paper and, I guess, allow
around 30 minutes of time.
Alex
A.P.Manners
[SNIP]
What you are doing has been widely done before as I am sure you realize.
May I suggest a quick google on the following:
"method of characteristics" (often shortened to MOC)
"filling and emptying"
It was a popular subject in the 1970s for things like simulating
combustion engines. There is certainly large amounts of literature and
code available but most of it is going to predate the web.
However, a better approach may be to simply work with a "1D" CFD example
code from a compressible flow book or course and work with that. Several
current courses and text books use such code to help teach CFD/fluid
mechanics.
A filling and emptying code does not use the momentum equation directly
so you are getting this right.
You do not refer to characteristics but recognize a need to the cell
sizes "just so". I would suggest that a quick look at the method of
characteristics is going to tell you why and what to do.
Brent Cullimore
Is this for an educational project, or do you have a specific need?
Seems like a lot of work unless this is a learning exercise, and
you'll need to validate the results to trust them.
Our finite difference/volume code can handle such effects, but the key
is to restrict the time step small enough (since it is implicit).
Yours sounds more explicit: how are you choosing a time step? That
may be where the cell size dependence enters, because otherwise
(beyond a minimum) the results in our code is largely independent
of discretization.
Good luck,
Brent
<snipped article>
-----------------------------------------------
Brent Cullimore, br...@crtech.com
C&R Technologies, www.crtech.com
Thermal/fluid Software and Consulting
Hot engineering ... Cool software (tm)
-----------------------------------------------
Brent Cullimore
<snip>
>
>Todd Wasson
>---
>Performance Simulations
>Drag Racing and Top Speed Prediction
>Software
Aha! Is this for tuned exhausts? If so, I know there used to be a
commercial program for this, but it seems to have gone bust or at
least disappeared, so you either have to use a general purpose code
or (as you are doing) write your own.
Brent
voy...@ihpc.net
"Brent Cullimore" <br...@crtech.com> wrote in message
news:ier97u4d77hgkpeni...@4ax.com...
> J Todd
>
> Is this for an educational project, or do you have a specific need?
> Seems like a lot of work unless this is a learning exercise, and
> you'll need to validate the results to trust them.
>
> Our finite difference/volume code can handle such effects, but the key
> is to restrict the time step small enough (since it is implicit).
> Yours sounds more explicit: how are you choosing a time step? That
> may be where the cell size dependence enters, because otherwise
> (beyond a minimum) the results in our code is largely independent
> of discretization.
>
> Good luck,
>
> Brent
you, probably, asking not about time step explicitly, but about correlations
between time and space steps, correct? For any time steps there is so small
space step, the problem solution could be not reliable, especially if you
have bifurcation point for under sonic and supersonic flows.
Alex
JohnB
<voy...@ihpc.net> wrote in message news:3c753...@news1.prserv.net...
is at http://www.optimum-power.com/
Instead of doing a finite element solution why not use method of
characteristics. It's well understood and described for both liquids and
gases. There are plenty of text and web references to check out.
rgds
JohnB
Flowmaster International
J. Todd Wasson
>>Drag Racing and Top Speed Prediction
>>Software
>
>Aha! Is this for tuned exhausts?
lol.. It'll be for a new engine simulation, yes. I'm testing the waters
using a better approach than I've done in the past. This new approach's
purpose is to more accurately account for intake and exhaust length tuning.
First aim is for a PC video game, then, if it's good enough, there are other
possibilities too... *IF* it's good enough, of course :0)
> If so, I know there used to be a
>commercial program for this, but it seems to have gone bust or at
>least disappeared, so you either have to use a general purpose code
>or (as you are doing) write your own.
Thanks a lot, but I'll need to write my own code for all this anyway.
Todd Wasson
---
Performance Simulations
Drag Racing and Top Speed Prediction
Software
J. Todd Wasson
Yes, I'm aware of that :0)
>May I suggest a quick google on the following:
>
>"method of characteristics" (often shortened to MOC)
>
>"filling and emptying"
If I'm not mistaken, the MOC has been surpassed by finite difference methods
by now, hasn't it? It seemed to me that MOC was used before fast computers
were commonly available. I'm no expert on that of course, having never used or
studied the MOC myself :0)
I've read about filling and emptying models, and started out (years ago, the
last time I attempted this type of thing) trying to make one of these, but the
problem with a filling and emptying model is you don't get any momentum effects
at all in the engine (as you said). I'm wanting to produce a full torque/power
curve here, and think an unsteady gas type model would produce better results.
Forgive me if my terminology is a little off, hopefully you know what I mean
(and can correct me where I'm mistaken).
[filling and emptying]
>It was a popular subject in the 1970s for things like simulating
>combustion engines. There is certainly large amounts of literature and
>code available but most of it is going to predate the web.
I've got Heywood's book, "Internal Combustion Engine Fundamentals", but once
the partial differential equations start popping up, I get lost quickly. All I
know about calculus is basically what a derivitive and an integral are, like
first chapter knowledge :0) Besides, for computer code, it needs to be
rewritten in algabraic form anyway as I'm sure you already know.
>However, a better approach may be to simply work with a "1D" CFD example
>code from a compressible flow book or course and work with that. Several
>current courses and text books use such code to help teach CFD/fluid
>mechanics.
Yes, this is really probably what I need :0) I'm talking now with a
professor of aerodynamics at Embry Riddle that's done his own CFD codes. He's
pointing me at a dissertation he wrote on the subject, so perhaps together we
can get my brain in gear regarding 1-D gas dynamics (he refers to "shock tubes"
for examples.)
>You do not refer to characteristics but recognize a need to the cell
>sizes "just so". I would suggest that a quick look at the method of
>characteristics is going to tell you why and what to do.
Ok, I'll search online and see what this is all about. Thank you very much!
My lack of knowledge in calculus is certainly an obstacle, I'm finding :0)
J. Todd Wasson
>is at http://www.optimum-power.com/
>Instead of doing a finite element solution why not use method of
>characteristics. It's well understood and described for both liquids and
>gases. There are plenty of text and web references to check out.
>rgds
>JohnB
>Flowmaster International
You're from Flowmaster, as in Flowmaster exhaust?
Thanks for the tips, I'll look into the method of characteristics as well.
J. Todd Wasson
>Seems like a lot of work unless this is a learning exercise, and
>you'll need to validate the results to trust them.
I'm going to see if I can just get the 1-D stuff working, then once the
"full" engine model is complete, I can validate it with a fellow that owns a
dyno shop. He's agreed to provide extensive engine test data for this purpose
(in exchange for a free copy of the program). I'm pretty confident that if the
1-D stuff works, the rest of the engine model can be made to get within a few
percent of reality in torque curve prediction. As it is, my QuickEngine
Builder is pretty good considering the assumptions it makes. Coding multiple
interconnected pipes will be a real b*tch though lol!
>Our finite difference/volume code can handle such effects, but the key
>is to restrict the time step small enough (since it is implicit).
>Yours sounds more explicit: how are you choosing a time step?
I was choosing a time step small enough to let me watch the pressure wave
move slowly on my 333mhz system :0) Otherwise, the time step is something I
was planning on just tinkering with, or possibly make self-adjusting.
>That
>may be where the cell size dependence enters, because otherwise
>(beyond a minimum) the results in our code is largely independent
>of discretization.
What does this mean? I'm not familiar with the term "discretization".
Thanks,
Brent Cullimore
I can't post binaries here or I'd show you a comparison we did with
MOC versus finite volume/difference that has less than 1% difference.
So you can get there either way. Finite volume/difference is harder
to write but much more versatile and powerful than MOC. MOC is
very easy to write (witness the hundreds of codes available ... many
written as student projects!) so that may be the best way for you to
go as long as all you care about are pressure waves in a fixed or
at least limited geometry.
Good luck, and let us know when we can beta test the game!
Brent
PS: JohnB, thanks for the URL. That's not the one I had been thinking
of, though. It was specifically meant for tuned exhausts for racing
machines. Can't recall the name unfortunately.
-----------------------------------------------
J. Todd Wasson
>I can't post binaries here or I'd show you a comparison we did with
>MOC versus finite volume/difference that has less than 1% difference.
>
>So you can get there either way. Finite volume/difference is harder
>to write but much more versatile and powerful than MOC. MOC is
>very easy to write (witness the hundreds of codes available ... many
>written as student projects!) so that may be the best way for you to
>go as long as all you care about are pressure waves in a fixed or
>at least limited geometry.
>
>Good luck, and let us know when we can beta test the game!
>
>Brent
>
>PS: JohnB, thanks for the URL. That's not the one I had been thinking
>of, though. It was specifically meant for tuned exhausts for racing
>machines. Can't recall the name unfortunately.
Thanks a lot, Brent. I'm looking now at MOC, and unfortunately, it looks a
bit over my head! lol
I had hoped that this could be roughly done by creating a bunch of constant
volume/area cells to make up an exhaust/intake runner. Then, by calculating
the approximate pressure in each cell, the force at each cell boundary could be
computed. I was hoping this could be merely added to each boundary's momentum,
or mass flow rate, directly, to get a simple approximation for now.
Unfortunately, the pressure waves don't travel at the correct speed, although
they do interact with each other appropriately :-( I must really be
oversimplifying the situation, I guess..
Todd
John Borgman
and static.
Every day is a present just waiting to be opened and explored
John Borgman
dynamics before you can even begin to claculate anything.
Go to referance section of your local library and get some basic info
that is neded for you to cmplete your project. Look for data on
laminar, transition and turbulant regions in flow. Also you need to
understand what the affects are on static pressure within a " tube "
carring airflow.
Do you understand that ot only can a liited amout of air be pushed
down a limited sectio of pipe? Know that with 7" pipe pushing 150 cfm
at.1 static, will have no air flow out the end if the run is far enough?
Hope this helps you with your project.
J. Todd Wasson
>dynamics before you can even begin to claculate anything.
> Go to referance section of your local library and get some basic info
>that is neded for you to cmplete your project. Look for data on
>laminar, transition and turbulant regions in flow. Also you need to
>understand what the affects are on static pressure within a " tube "
>carring airflow.
Thank you for your reply, John. However, I after reading part of a
dissertation on the subject of shock tube simulation, I found a small error.
By dividing the momentum change across each of the boundaries by the distance
between cells, the problem was solved, and the pressure waves began moving at
the sound speed. So.... I'm inclined now to believe that air essentially does
work this way after all :0)
> Do you understand that ot only can a liited amout of air be pushed
>down a limited sectio of pipe? Know that with 7" pipe pushing 150 cfm
>at.1 static, will have no air flow out the end if the run is far enough?
> Hope this helps you with your project.
I don't understand exactly what you mean here...?
With a 7" pipe flowing 150 cfm measured at one end, there will be no airflow
into the other end? I must be misunderstanding your statement.
Mike
>
> With a 7" pipe flowing 150 cfm measured at one end, there will be no
airflow
> into the other end? I must be misunderstanding your statement.
>
>
Yes, I think he is mistaken. If you apply a pressure to a long length of
pipe, you may get zero flow out of it if the pressure drop of the pipe is
substantial enough.
However, if you force fluid through at a given flow rate, you must get flow
out of the other end; pressure would build and therefore overcome wall
friction.
J. Todd Wasson
>pipe, you may get zero flow out of it if the pressure drop of the pipe is
>substantial enough.
>
>However, if you force fluid through at a given flow rate, you must get flow
>out of the other end; pressure would build and therefore overcome wall
>friction.
>
>
Oh, ok. Thanks for clearing that up. It's good to hear validation that I'm
not nuts once in awhile :0) The scenario I was picturing was a pipe open to
the atmosphere on one end, filled at atm pressure, with the other end suddenly
opened to a lower constant pressure volume. My thinking is, no matter what,
there's going to be flow from the pipe out to the low pressure zone across that
boundary, regardless of the length of the pipe. By definition in this system,
there's a constant pressure drop across the pipe ends, so this doesn't really
say how much flow there would be, but just that there would be some flow.
What I'm finding from my cheezy model experiment is, the mass flow rate
through this low pressure end will rise quickly to a certain level, then stay
pretty constant while a pressure wave travels to the other end of the pipe and
back, then jump pretty quickly to a higher level where it more or less
stabilizes, then again and again, until friction/flow restrictions dampen it
out to a steady state condition throughout the pipe. Am I on the right track
here? And this pressure wave travels at the sound speed relative to the local
flow velocity in each of my "slices" in the pipe (which itself would constantly
be changing and different), right ?
I've read that a compression wave steepens and a rarefraction wave relaxes as
it moves. Eventually a compression wave turns into a shock as it reaches
infinite "steepness" (word?) if the pipe is long enough (and friction is
neglected, probably). However, I haven't checked to see if this actually
happens or not in my model yet. I have a feeling that once I tweak one other
thing, it'll happen though, but probably will get messed up once the shock
itself occurs. Then, I'm back to trying to get Navier-Stokes working again..
I've got some of that coded, (I think :0P), but of course, it's not working
because I'm missing something.
Anyway, I'm rambling now.
Thanks for any thoughts,
David Wilkinson
thoughts but your experiment is, of course, giving the right answer. You
could look up water hammer in Fluid Mechanics books for a more complete
description.
For your pipe filled with atmospheric pressure air with one end suddenly
dropped to a lower pressure, a rarefaction wave will start at the low
pressure end and travel at the speed of sound towards the other end. The
pressure drop across the wave will be deltaP, the amount you lower the
pressure. There will be a corresponding velocity change across the wave
of DeltaV = DeltaP/Rho/c where Rho is the density and c is the wave
speed or speed of sound.
Assuming you are in the USA and use Imperial Units, Rho = 0.002378
slugs/ft^3, c = 762 mph = 1118 ft/sec. Suppose you drop the pressure by
1 psi = 144 lb/ft^2 then the velocity change is
DeltaV = 144/0.002378/1118 = 54 ft/sec.
As soon as you drop the pressure air starts flowing out of the low
pressure end at 54 ft/sec. As the wave moves towards the high pressure
end then on its low pressure side the pressure is 13.7 psi and velocity
54 ft/sec. On the high pressure side it is still pressure = 14.7 psi,
velocity zero. This air does not know the low pressure end has been
opened until the pressure wave arrives.
When the rarefaction wave reaches the atmospheric end it can't drop the
whole atmosphere by 1 psi so the pressure there stays the same. It does
this by the wave reflecting with change of sign to a 1 psi pressure wave
which travels back towards the low pressure end, raising the velocity to
108 ft/sec as it passes.
The wave keeps reflecting from the open ends with change of sign until
friction intervenes and it reaches a steady state.
In article <20020312163808...@mb-mh.aol.com>, J. Todd Wasson
<jtw6...@aol.com> writes
--
David Wilkinson
J. Todd Wasson
>thoughts but your experiment is, of course, giving the right answer. You
>could look up water hammer in Fluid Mechanics books for a more complete
>description.
>
Oh, ok.. I will make this distinction in the future. I read about water
hammer some time ago and implemented some sample code from the book. Although
it was educational, it wasn't quite what I was looking for. My current stuff
works much better for what I'm doing. Thank you for the suggestion though.
Also, thank you very much for the analysis you gave here! That's exactly what
I need. I'll compare the numbers to make sure it's actually working right, but
that describes exactly what the flow in my tube did and explains how and why
the mass flow rate (or more specifically it turns out, thanks to you, the
velocity) changes instantly to what it does. This was a confusing thing to me
for a long time until I started this "new" (for me anyway) method, which is
supporting this quite well. Thanks for clearing that up and giving me
something to test the model against.
I'll let you know (if you care to, that is ;-)) how the numbers work out.
Also, one more question if you don't mind. This pressure wave that moves up
and down the pipe.. Each time it traverses the pipe, the flow velocity
increases (ignoring friction). Does this pressure wave move at sonic speed in
relation to the flow itself, or the pipe?
I saw an eigenvector something like the following used for corrective flux in
the Navier-Stokes equations in the event of a shock wave (maybe it's used
elsewhere too, don't know yet):
a-u
u
u+a
I didn't really look to see where it's specifically used, but I've taken it as
an indication that the pressure wave moves at sonic speed relative to the gas
velocity. This would mean that once the fluid flow itself reached Mach 1, the
pressure wave would stand still... True? I don't know that I'm too concerned
really with shock waves in the pipe systems, unless it was for a top fuel or
alcohol exhaust system or something, but while we're at it, I'm curious how
that works (those guys have their own engineers for this type of thing anyway,
so... :0))
Maybe that's the point where this model would fall apart no matter what. If
so, that's probably fine, as this is extremely fast and efficient as it is
right now (compared to other experiments). But then again, maybe it's right;
if the pressure wave stood still once one end reached sonic speed, that would
prevent the wave from ever returning again to increase flow velocity above Mach
1, which seems to satisfy at least one condition for choked flow... But, on
the third hand, it seems that a shock wave is an infinite steepness compression
wave that travels at Mach 1, which means this is not true....
Where am I lost?
Thanks again,
J. Todd Wasson
>if the pressure wave stood still once one end reached sonic speed, that would
>prevent the wave from ever returning again to increase flow velocity above
>Mach
>1, which seems to satisfy at least one condition for choked flow...
But, on
>the third hand, it seems that a shock wave is an infinite steepness
>compression
>wave that travels at Mach 1, which means this is not true....
>
Oh wait.. If the shockwave travelled at Mach 1 relative to the fluid flow,
which was itself moving at that speed, the wave would stand still.
David Wilkinson
<jtw6...@aol.com> writes
depending on which way it is going. My very simple analysis was for
small pressure and velocity changes, usually called Mach waves. If you
are into velocities of more than 0.2 to 0.3 of the wave speed then you
need the full compressible flow treatment and may have large shock waves
that move at a higher speed than the speed of sound.
If you push hard enough the wave can move at any speed relative to the
fluid. Think of a space vehicle re-entering the Earth's atmosphere. It
has to get up to about 25,000 mph to get out of the gravitational field
completely so it must be going at this sort of speed on re-entry. That
would be about Mach 33 at sea level but must be rather more in the
colder upper atmosphere. So, you can push a normal shock along at more
than Mach 33, showing the wave never stands still. The speed of sound is
the minimum wave speed for weak shock waves or Mach waves.
>Maybe that's the point where this model would fall apart no matter what. If
>so, that's probably fine, as this is extremely fast and efficient as it is
>right now (compared to other experiments). But then again, maybe it's right;
>if the pressure wave stood still once one end reached sonic speed, that would
>prevent the wave from ever returning again to increase flow velocity above Mach
>1, which seems to satisfy at least one condition for choked flow... But, on
>the third hand, it seems that a shock wave is an infinite steepness compression
>wave that travels at Mach 1, which means this is not true....
>
If the pressure ratio across the pipe gets above about 1/0.5283 = 1.893
for air in a steady state, the flow will choke, either at the minimum
area or at the downstream end if the duct is constant area. This means
the Mach number in the throat will be one and the flow in the duct will
not change even if you reduce the downstream pressure further. I have no
idea what pressure ratios and Mach Numbers are reached in engine inlet
and exhaust ducts. If the flow chokes at an internal throat and there is
then an expansion in area the flow will expand into the supersonic
region, possibly shocking down again to subsonic before the exit
depending on the downstream pressure. You may need to read a
compressible flow book.
>Where am I lost?
>
>Thanks again,
>
>
>Todd Wasson
>---
>Performance Simulations
>Drag Racing and Top Speed Prediction
>Software
>http://PerformanceSimulations.Com
>
>My little car sim screenshots:
>http://performancesimulations.com/scnshot4.htm
>
--
David Wilkinson
ATREAS Consulting
I have just browsed through this thread and realized that
most of the responses just show ignorance and half-knowledge
of a subject that it is not as simple as it sounds.
For more info on this and other CFD matters, visit
our site at
ATREAS Consulting
-----------------
David Wilkinson <da...@quarksoft.demon.co.uk> wrote in message news:<7mQMUJAa...@quarksoft.demon.co.uk>...
A.P.Manners
>
> Hi all!
>
> I have just browsed through this thread and realized that
> most of the responses just show ignorance and half-knowledge
> of a subject that it is not as simple as it sounds.
>
> For more info on this and other CFD matters, visit
> our site at
>
> www.atreas.gr
>
> ATREAS Consulting
Am I interested in the services of this company if the author:
* has not made a helpful contribution to the thread
* has not demonstrated a grasp of the subject
* has blatantly plugged a commercial service out of context
* has pointed at a web page which is an almost unreadable mess in my web
browser
I continue to be amazed that employees of apparently serious companies
(although one can never be that sure on the web) damage the image of
their companies by posts such as this one. Would it have been that
difficult to make a positive contribution to the thread and work in the
odd reference to the enterprise in a way that does not annoy people?
/*/
>
> Hi all!
>
> I have just browsed through this thread and realized that
> most of the responses just show ignorance and half-knowledge
> of a subject that it is not as simple as it sounds.
>
> For more info on this and other CFD matters, visit
> our site at
>
> www.atreas.gr
>
> ATREAS Consulting
wow! whatever the merits or otherwise
of the technical argument, your marketing
skills are nil. or, is it that being
obnoxious is part of your sales pitch?
--
J. Todd Wasson
>I continue to be amazed that employees of apparently serious companies
>(although one can never be that sure on the web) damage the image of
>their companies by posts such as this one. Would it have been that
>difficult to make a positive contribution to the thread and work in the
>odd reference to the enterprise in a way that does not annoy people?
>
>
No kidding.. This guy emailed me too with the same thing, only didn't even
bother to provide a link to his site (hotmail address that I was not to respond
too). In this thread, I got exactly the answers I needed, and after doing my
stuff, it works great thanks to you guys.. I didn't get any "crap" out of this
place at all... until now :-)
ATREAS Consulting
>
> Am I interested in the services of this company if the author:
>
> * has not made a helpful contribution to the thread
The thread is old. No helpful contribution could be made
at this point. Furthermore, the services of ATREAS Consulting
are not free.
>
> * has not demonstrated a grasp of the subject
This is obvious in the web site.
>
> * has blatantly plugged a commercial service out of context
One of the first responses to the problem stated :
---
"What is 1-d pipe? I think pipe has 2-D geometry. What means "pipe" for you?"
---
I think, that in this context a visit to our site is quite informative.
This newsgroup is free to anyone. However I don't think that the
above statement has been written from someone with any grasp of CFD.
>
> * has pointed at a web page which is an almost unreadable mess in my web
> browser
Netscape is old. Update it or use IE.
>
> I continue to be amazed that employees of apparently serious companies
> (although one can never be that sure on the web) damage the image of
> their companies by posts such as this one. Would it have been that
> difficult to make a positive contribution to the thread and work in the
> odd reference to the enterprise in a way that does not annoy people?
You should not be annoyed, since you know that your contribution
has been positive and right in the context. The problem was contributions
such as the one I mentioned above and these two :
---
"When you do fluid flow along axis in a pipe, you have Navie-Stoks, which is
orthogonal of speed of you media, that create parabolic profile of media
velocity. I see 2D problem in your description. Does it correct?"
---
In that case the contributor does not know anything about parabolic,
eliptic or HYPERBOLIC flows (as in this case) and speaks about 2D
modelling !
Another great contribution was :
---
"P.S. It is impossible for fluid or air to flow without swirls, eddies
and static."
---
Do you really find that these contributions were positive and helpful to
someone without any Fluid Dynamics or CFD Background ?
To our view, it doesn't make sense to try to help at "sci.medicine", since
we have nothing to do with medicine. I hope that this made sense to you...
ATREAS Consulting
ATREAS Consulting
> No kidding.. This guy emailed me too with the same thing, only didn't even
> bother to provide a link to his site (hotmail address that I was not to respond
> too).
We keep copies of our emails. The link was there.
> In this thread, I got exactly the answers I needed, and after doing my
> stuff, it works great thanks to you guys.. I didn't get any "crap" out of this
> place at all... until now :-)
We didn't said that all contributions were negative. However we considered
some contributions were unacceptable (see previous posting).
Furthermore we think that since everybody has the right to post
information and "help" of the quality we saw in those postings,
we also have the right to criticize them.
ATREAS Consulting
ATREAS Consulting
> > our site at
> >
> > www.atreas.gr
> >
> > ATREAS Consulting
>
> wow! whatever the merits or otherwise
> of the technical argument, your marketing
> skills are nil. or, is it that being
> obnoxious is part of your sales pitch?
> --
It is! Would you have visited our site otherwise ? ...
ATREAS Consulting
>
>
David Wilkinson
Consulting <atre...@hotmail.com> writes
words I would use. As a sales pitch on a scale of 0 to 10 it gets about
minus 3. If your self-proclaimed expertise matches your PR then I would
rather not know about it.
--
David Wilkinson
J. Clarke
atre...@hotmail.com says...
Visit <http://www.stonebrewing.com>. Their major product should be your
favorite beverage. Not that you're worthy.
--
--
--John
Reply to jclarke at ae tee tee global dot net
(used to be jclarke at eye bee em dot net)
/*/
tritokosmiki thriambologia, blaka!
aei xasou.
--
>
> >
> >
Andy
OK so you have responded which is a sufficiently good sign for me to
reply. But...
> > * has not made a helpful contribution to the thread
>
> The thread is old. No helpful contribution could be made
> at this point. Furthermore, the services of ATREAS Consulting
> are not free.
This is not a justification for posting but quite the opposite given
that contributors to newsgroups are required not to post advertising.
> > * has not demonstrated a grasp of the subject
>
> This is obvious in the web site.
Not really. The web site is a promotion which may indicate what areas
you are willing to work in but does not demonstrate knowledge in the
way taking part in a debate will.
Also, the site lacks technical details which would have strengthened
you case. In addition, the methods you are promoting are not current
and most, if not all, could be picked up elsewhere for the cost of
integrating them into one's design procedures. This is not a severe
criticism since you are presumably a young company and willing to work
for rates significantly less than an established company. Again, I
would suggest trying to generate goodwill is a wise move given your
intended line of business.
> > * has blatantly plugged a commercial service out of context
>
> One of the first responses to the problem stated :
[SNIP]
> I think, that in this context a visit to our site is quite informative.
> This newsgroup is free to anyone. However I don't think that the
> above statement has been written from someone with any grasp of CFD.
I cannot see how visiting your web site would have answered/clarified
this specific question. If it could, then a quick explanation how/why
and a link would be a positive contribution which few would object to.
> > * has pointed at a web page which is an almost unreadable mess in my web
> > browser
>
> Netscape is old. Update it or use IE.
Now you are betraying ignorance concerning CFD users and their
computational platforms. Most serious CFD users do not use Windows but
a flavour of unix. IE runs on almost none of these platforms and
Netscape is by far the most commonly used browser. I updated to the
latest release of the browser on my platform less than 2 months ago.
If you wish to learn how to develop standard conforming and widely
viewable web pages then all the information you need is at
www.w3c.org.
Again, your company is sending out the wrong signals to your potential
customers. That is, if you don't use Microsoft products we are not
interested. (I am assuming that you are actually targeting non-MS
customers but I could, of course, be wrong).
> > I continue to be amazed that employees of apparently serious companies
> > (although one can never be that sure on the web) damage the image of
> > their companies by posts such as this one. Would it have been that
> > difficult to make a positive contribution to the thread and work in the
> > odd reference to the enterprise in a way that does not annoy people?
>
> You should not be annoyed, since you know that your contribution
> has been positive and right in the context.
I am not annoyed about my contribution but yours. (Although annoyed is
too strong - it is more an irritation).
> The problem was contributions
> such as the one I mentioned above and these two :
[SNIP]
Personally I do not consider these a problem because the person is
trying to help. So long as there are only a modest number of such
postings the newsgroup does not really suffer and, often, a correction
would be generated by someone else benefiting (at least) two people.
What is required is tolerance. Tolerance of people being wrong in
order to generate discussion and tolerance of people, such as
yourself, with something to sell so long as they contribute
positively.
> Do you really find that these contributions were positive and helpful to
> someone without any Fluid Dynamics or CFD Background ?
Indirectly yes because they were part of a discussion. Directly no
because the information contained was not useful.
> To our view, it doesn't make sense to try to help at "sci.medicine", since
> we have nothing to do with medicine.
I hope "our" is not supposed to include myself. I am sure there a few
instances where I could make a helpful contribution to sci.medicine
and would happily do so.
> I hope that this made sense to you...
Yes, I think I understand what you are trying to convey but I do not
agree with much of it.
> ATREAS Consulting
At the risk of passing yet more unwanted advice, if you wish to build
goodwill using your own name would be a (small) improvement. Anonymous
(or partially anoymous) postings are not reponded to in quite the same
way as those from identifiable people.
J. Todd Wasson
>even
>> bother to provide a link to his site (hotmail address that I was not to
>respond
>> too).
>
> We keep copies of our emails. The link was there.
>
>
Ah, yes, it was indeed there now that I look again. I apologize for my snide
remark.
>We didn't said that all contributions were negative. However we considered
>some contributions were unacceptable (see previous posting).
>
>Furthermore we think that since everybody has the right to post
>information and "help" of the quality we saw in those postings,
>we also have the right to criticize them.
Yes, I agree with you here, you do have that right.