Gordon,
>Kit, thanks for your patience on this one. I have included a few published
quotes below which may help clarify... <
No problem, and thanks for looking them up.
>...solving all of the trailer car issues meticulously identified by the
Ford/Dain review of the APT. The descision to ditch a proven, >lighter
(higher performance) solution, in favour of a new Mk 4, seems bizarre from
a technical standpoint. <
Who said technicalities had anything to do with it? <g>
I suspect it was a political decision, driven by marketing men, the object
of the exercise being TOTAL rebuttal of anything even vaguely connected
with APT.
>Furthermore, there is no mention in the study of real data from the APT.
They simulated a Mk 3, and compared to real data, but the >aluminium
designs were only simulated. <
You can make anything work, or not work, in a simulation. It depends who's
asking for the work to be done and what they REALLY want to know. With a
sim. you can easily produce data that proves any point you want to make,
it's like market research.
>I also found that a Railway Gazette article from 1980 on APT stated that:
>"Structural stiffness is such that the lowest natural frequency of
vibration of the vehicle bodies is high enough to avoid coupling with >the
suspension frequencies. Intermediate trailer cars have minumum vertical and
lateral bending frequencies of 13.2Hz and 15.3Hz >when full equiped."<
Makes some sort of sense, yes. (see Lesson 1 <g>)
>This all seems above reproach, and it's strange indeed that the aluminium
bodies in the study came out at 8.2Hz. Is it possible that >dearticulating
could have made such a big difference? <
No, it's not relevent to the Primary modes. Articulated bodies ought to
show lower primary frequencies than a non-articulated ones, as the distance
between the support points is possibly greater, and thus the nodes are
further apart too, but not THAT much lower!. That's assuming a similar
structure of course.
>Or could it have been the desire reprofile the body to exploit the, by
then reduced tilt angle, to give more space inside? Or could it be >the 2m
length difference? Or had BR already started to loose the skills required
to tweak the APT structure?<
Mk 4s were taller than, P-Train weren't they? That ought to give greater
beam strength and thus higher Primary Bending frequencies, but it depends a
lot on how the rest of the structure was tied together. I could tell you
about the Prototype High Density Stock and it's sliding doors...... or
non-sliding, as the case maye be!
And which one was longer? 2m would make a difference if it was echoed in
the node pionts, but I doubt they'd be that different as the nodes would be
different distances from the vehicle ends in each case, less distance on
the P-Train coach.
The last point may be relevant, people were leaving in droves at that time.
>They used finite element analysis of the Mk 4 structure in response to
real track stimulus, modelling 63 degrees of freedom. <
SIXTY THREE? Ye gods, that's gilding the lily a trifle, or maybe clutching
at straws?
>The APT paper which Boocock and King presented to the IME in 1982 may shed
some light here:
>"One aspect of ride which has proved difficult to satisfy by improved
suspension design is related to long (above 40m) track defects. >These are
beyond detection and correction by present track maintenance methods. As
APT-P suspension frequencies correspond >to wavelengths of 45-80m at
200km/h, low frequency oscilations can be pronounced on certain stretches
of track. <
That says the first fixed body bounce mode was about 1.0 Hz or so. 45 m at
that speed equivalent to 0.8 Hz and 80 m to 1.44 Hz. Pretty uncomfortable
I'd have thought.
>Vertical ride development has been concentrated mainly on the articulated
bogie. To compensate for the outboard positioning of the >secondary
suspension systems, the airsprings were designed to be very soft, with
auxiliary surge resorvoirs to provide damping. >Nevertheless, the vertical
ride was found to be unsatisfactory, even after tuning the air suspensions
and increasing their volume. <
I'm not that good on pnuematics but if that was a hydraulic system we'd
have put damping orifices in the line between the springs and the
reservoir. I did some tests on exactly that for E-Train in the Lab some
years before. I wonder if they had the dampers fitted before Run One?
>These problems were solved by fitting vertical hydraulic dampers between
the ends of adjacent vehicles. These inter-vehicle >dampers also slightly
improved the ride about the roll axis."<
Makes sense but it's only curing the symptom, not the illness. Of course
the dampers couldn't be fitted on the centre-line as the gangways were
there, so being offset they would have quite marked effects in the roll
axis.
>Three structures were examined, a Mk3 type steel structure, an aluminium
structure, and aluminium body with a stiffening keel below >the passenger
compartment as on the prototype Advanced Passenger Train (APT-P). While the
deep keel is useful structuraly, the >physical barrier complicates the
instalation and maintenance of the underfloor equipment. <
Cobblers...............
It may make the design of the equipment difficult as it is limited in width
to less than 50% of the body, but it gives a good bulkhead to attach all
the fixed bits too, plus it helps keep the dirt out, always a problem on
railway vehicles. Brake dust gets EVERYwhere! Of course, the fact that
P-Train had a keel and the Mk4 wasn't going to has much bearing on this.
See my comment about total rebuttal of anything to do with APT.
Our Mk 4 tilt packs, as fitted to POP Mk 2 and Pilot, and later production
versions to P-Train, fitted nicely in that half underbody. The control
cables and hoses were bulkheaded to the keel and the pack hung from the
floor. Worked a treat and was relatively easy to remove and replace, as it
was designed to. I suspect the above comment is concerned with doing the
maintennace on the vehicle, and P-Train was designed to do exactly the
opposite, everything was meant to be hauled out and replaced with
pre-checked modules. Of course that meant the whole idea was flawed to
their thinking and would not be considered for Mk4s or anything else.
>While significantly lighter at just over 6.5 tons, both aluminium shells
were very close to the minimum frequency , the best being the >keeled
version at 8.2Hz.<
Yeah, right. Where's the real test data on real structures to prove it? I
know it exists, we did the tests in the APD Lab on Test Coach
Pilot........
I just don't believe that a P-Train bodyshell was that low in Primary
Bending. That also purports to say that a non-keeled version was even
LOWER, for goodness sake! What was it made of, toffee?
It all sounds to me as if they were running 'tests' to prove a point they'd
already decided on for other reasons. Me, a conspiracy theorist? How did
you guess? <BG>
Very interesting stuff nonetheless. I wonder what did happen to those
vibration test results on Pilot though?
Regards
Kit