Pull through or just let it rip?

What they're testing for is not relevant at all. I'm simply responding to your original assertion/suggestion that B&M is so confident in their engineering that they don't have to test for things. I wasn't referring to any specific type of testing - in fact, I was assuming the pull through being inquired about was in regard to simply making sure the train travels along the track as intended.

Basically, all I'm saying is that B&M having the "confidence" to not do a pull-through isn't something I'd want my engineering firm to be known for. Testing isn't a sign of weakness or poor engineering skills. In fact, quite the opposite.

I said they don't have to test for clearance envelope, not "things." They've virtually tested, so why would they need to do it physically?

Because virtual testing isn't real?

Its the same reason that auto manufacturers (yes, even Hondoyota) design, build and then test prototypes in the real world. There's no amount of virtual testing and/or calculation that can make up for simply building the damn thing and trying it out under controlled conditions. Not doing so under the guise of we're so effin' smart, we don't need to check our work flies in the face of the most basic engineering principles.

ADDED:
You tested your forum app, right?So, why not simply code it and release it, consequences be damned? I mean, you're smart, so it should just work! :)

This isn't a binary decision tree we're talking about here. You can't be dogmatic about it just because you want to make a point. Even in software development, it's rare that you test every line of code because the risk to cost ratio makes it not worth it. B&M rides are already more expensive than their competitors, and they believe "throwing it over" is a low risk thing to do. They'd be on the line for million dollar trains, so it seems to me that wouldn't be a decision they'd take lightly.

So, B&M does not pull trains through any of their coasters? I was assuming that you were speaking speculatively, though you seem certain. Just curious.

And as for your analogy to testing every line of code, c'mon. You know there's a massive difference between testing every line of code and making sure the thing functions as intended. That is, does the forum work? Similarly, they wouldn't test every single nut, bolt and thingamajig on the coaster - but they'd sure make sure the thing functions as intended on its most basic level. That is, a train can complete the course without binding or catching on something or whatever.

Perhaps, as you assert, they figure it would cost them X numbers of days to do a pull-through, and because a train "only" costs a million dollars, their time is worth more than that. I can't imagine that to be the case, unless they're using UAW workers. ;)

djDaemon said:
That is, a train can complete the course without binding or catching on something or whatever.

Doesn't throwing it over the hill and seeing it come back constitute a successful test?

Of course it does. At significant potential cost in the event of that "test" failing.

Why don't Toyota engineers just design a vehicle and dump it into showrooms right away? If they've done the math up front, everything should be fine.

You can compare it to whatever you want all day long (as you seem intent on doing), it doesn't make it the same thing. Roller coasters, and especially B&M rides, are not complicated machines.

I think some people are missing the point.
Yes, when Toyota comes up with a new model-they test 100's of them in every way possible.

They open and close the doors 22,000 times.
The kick the tires 68,000 times.
They even do crash testing.

But after these "Test" cars are approved.
They start mass producing the ones you will actually Buy.
These only get a few quick inspections.

No one wants to buy a car that has 20,000 miles on it as new.

B&M is no different.
All of the test have been done, in the past- now they are into Mass Production.

All that a pull through does is this-It insures that the clearance between the two rails is correct.
That there is not enough tolerance for the train to either- come out of the tracks,or bind between them.

Considering the fact that each section of track is built in a shop ,and measured there, Has this test not already been done?

And if Everything Bolts together correctly, then it would reason that the footers are placed correctly.

Just like with the Toyota-"who wants to buy a coaster that has been worn out through multiple testing?"

I understand that we are talking about rollercoasters here, but I would like to give another example. The Citigroup Center in New York was completed in 1977 and was designed by a well respected engineer William LeMessurier. As with all skyscrapers this building was tested in a wind tunnel hundreds of times and all checked out ok. But it was brought to LeMessurier's attention that all was not ok with the building, and in fact a large oversight had been made in the building of the structure, which caused emergency repairs to be done immediatly. The reason I have recounded this, is that no matter how much "virtual" testing is done on any structure from rollercoasters to skyscrapers it is plain idiocy to take virtual test numbers and data over real world numbers and data.

djDaemon said:
Similarly, they wouldn't test every single nut, bolt and thingamajig on the coaster - but they'd sure make sure the thing functions as intended on its most basic level. That is, a train can complete the course without binding or catching on something or whatever.

On the contrary, every single nut, bolt, etc. does get tested.

The train completing the course also gets tested... by pressing "dispatch" and waiting for it to return.

Said by a guy who has likely pressed the button. Is that good enough?

You know, somehow I don't think they had the computing power to virtually test complex structures in 1977. I see to recall digital calculators being a big deal at the time, and ridiculously expensive.

Then again it might be B&M doesn't need to do pull throughs since they use the same cookie cutter elements and layouts over and over again. Like said B&Ms are in mass production theres no need for anymore test, compared to Intamin, with Intamin it's a different model pretty much everytime.

Also a computer can't can't stimulate everything, and when your pushing the envelope to the max like Intamin does then real world test is a must Imo. Like when was the last time B&M had to design a new support for a highly twisted , highly over banked ,or insanely steep element, like never. It's always the same supports, same elements, same boring lines layouts. The wildest thing B&M does is their Dive Machine and that's only a 90* drop and the same blah blah blah after that, sorry but B&M sucks when it comes to creativity so no need for a B&M pull through b/c theres nothing new imo .

B&Ms may seem cookie cutter, but the elements are of different sizes from one ride to the next for loopers, and no two hypers are the same size either (clones excluded).

Yeah, OK. B&M is teh suX0rz!!!11one!!! That adds a lot to the discussion of testing.

I think I remember seeing a picture of a powered test car with a camera that was sent around the track on Silver Bullet (or one of the other inverts) before they let a train fly. Anyone else know what I am talking about or was that something I imagined?

You might be thinking of the maintenance "pod" they put on Dominator/Batman Knight Flight at Geauga Dominion. I could be wrong, but I thought the reason for that was the relative difficulty involved in getting to places on the track to torque bolts and smooth out the seams between track sections, given that it was largely over water.

All of your analogies are completely flawed. Completely.

If Toyota puts out a car that is untested, and flawed because of a screw up in 'virtual' testing, then that car can and may kill people. Possibly large amounts, prior to being noticed.

How precisely do you test a building besides putting a model in a wind tunnel and other virtual testing? You build the damn thing and let it stand. The only way to find if there are flaws are by the virtual testing, someone actually noticing it, or catastrophe (and quite likely death). So what does a building and a screwup in one have to do with this?

Testing in software is much more like testing a roller coaster, yet again, not quite. You test major functions, if things fail - no one dies. You don't test every single process or component. Why? Because no one dies.

Roller coaster testing is the same, in that, no one dies. If B&M chooses to test by just letting the first train run on it's own, and something bad happens, no one dies. They put test equipment and dummies in the train to make sure (after that initial test run).... no one will die. So what is the worst that can happen by "letting her rip" and you all seem to like saying? Well, a train, and possibly some track get torn up, but again, no one dies as a result.

A roller coaster is checked and double checked. B&M is confident in their engineering ability so much so that they are not concerned about a train meeting its demise during testing.

djDeamon, well, I'd imagine B&M is not really all too concerned about what you'd want your engineering firm to be known for, now are they? Quite logically (with anyone who isn't biased), they're known for building solid rides that work day in and day out, meet opening day deadlines, and don't experience catastrophic failure on their first test run.

Can you cite one instance in the 70 or whatever roller coasters they've built where one failed to complete the course on it's first run and ruined an expensive train. I'd be interested to know of such a thing.

You're right Jeff. It's called the B&M Ball, a small self-powered maintenance pod designed to navigate around any B&M coaster track to help with inspection. There are only a handful of these things in the world and Kings Dominion has one (presumably from GL).

There is a picture of it here.

Heree is a picture of the maintenance pod....

http://rcdb.com/ig720.htm?picture=5