Cedar Point guest hit by Top Thrill Dragster part while in queue

About a total redesign...

Why not use cameras and AI to allow the computer to always know where the trains are? People have X-Box cameras in their living room that can map out the entire room and the people in it. Automobiles can practically drive themselves. Use Rise of the Resistance as an example, except with the technology and use for a roller coaster.

Or are we just not there yet?

Because it's Intamin? ;)

Red Force has the advantage that the motors are the brakes in the event of a rollback, but it seems silly that they interspersed the end of the run with retractable fins. Use all motors, which may cost more, but you save by not having an entire subsystem to raise and lower those fins and the control system to operate them.

HeyIsntThatRob? said:

OhioStater said:

That's what I was suggesting a page back. That type of change requires taking a step back and approaching the whole concept with a fresh approach.

This is a chance to (perhaps) rethink it. The whole thing.

See Red Force at Ferrari Land.

In that design the hydraulic system has been completely abandoned, but I don't know if anything changed on blocking and number of trains. I know they made some changes to King da ka based on what they learned on TTD.

I learned at WCO that there was some serious upgrades to the computer system and the hydraulics for the launch part of TTD. If I understand, the PLC system only controls positioning and blocking of trains outside of the launch. Then the launch part is its own system. There is some sort of handoff between the two that is done.

I think the window of starting with a clean sheet has passed based on scope of this last upgrade. Again, this info came from my conversation with maintenance, so it could be a telephone game. It seems some news articles confirm the upgrade that was done.

LSM's are not as efficient at accelerating trains as an hydraulic launch, so even though Red Force has the same launch track length as Top Thrill Dragster, Red Force accelerates its trains to 107-111 mph in 5 seconds. In addition, the ride only has 3 trains and each train has 3 cars, for 12 passengers, again to compensate for the LSM's.

Braking on Red Force is done a bit differently too: after dropping off the 367 feet tall spike, the train hits permanently mounted metal fins before rising up up into the final brake run. This make sense, as magnetic brakes generate a lot of heat and the brief section where the train is not trimmed as it rises allow for some heat to dissipate.

Intamin still technically offer the hydraulic launch, but haven't opened one since Formula Rossa at Ferrari World Abu Dhabi in 2010. The new generation LSM Coaster with the powerful Indrivetec LSM's (compared to other LSM systems) have completely replaced them.

I remember when Dragster was built there was discussion about why they went with a hydraulic catapult instead of an electromagnetic one, and the answer at the time was that the LSMs were just not powerful enough yet to accelerate the train in the space allotted. How much longer would it need to be? I haven't a clue. But then, could they have done something more like Volcano...put the station under the tower and launch the other way into a giant horseshoe turnaround, into a second launch to go even faster?

As for detecting the trains...are there other ways to do it? Certainly. Heck, Gemini uses optical pairs and cat's whisker limit switches for its block detection. It also can't run in the rain, and has problems at certain times of the day and season due to sunlight reflecting off of the bumpers and blinding the optical detectors. Or if you really want to be precise about it, how about a timing comb like Wicked Twister uses? There are lots of options; in this case they used something relatively inexpensive, simple, reliable, and for almost 20 years, safe.

--Dave Althoff, Jr.

I wouldn't consider anything with hundreds of sensors to be simple, and reliable is not a word associated with that ride. Just saying.

Jeff said:

I wouldn't consider anything with hundreds of sensors to be simple, and reliable is not a word associated with that ride. Just saying.

Every time you have a ride with multiple trains, especially one that requires to move 2 or 3 trains to move at the same time in the same blocks, you need a lot of sensors to know where they are. Space Mountain at Disneyland and Tokyo Disneyland is a fascinating and unique example: it can run 11-12 trains, dispatched on a 20-22 seconds interval. How is that possible? In the station area, pneumatic cylinders swiftly push rockets ahead, as soon as the previous rocket clear the middle of its stop position. You can imagine how many sensors are present in each block to keep track of the rockets as they move in an amazing ballet.

Top Thrill Dragster appears similar, using traditional pusher tires, but this was simplified over at Kingda Ka. Kingda Ka originally used two stations, each with two positions for trains. Eventually, it was discovered that going to a more standard set-up where running three trains on one loading position was as efficient as running all four trains in all loading positions.

Rideman, I'd estimate you'd need to knock down Gemini and go all the way to the perimeter road to get above 400 feet for the top hat with LSM's. If CP instead retrofit LSM's on the existing launch track, they'd need to lower the top hat to Red Force height, 367 feet tall and go to 3-4 car trains.

The sensor by itself is inexpensive, simple, and reliable. The complexity comes in with having lots and lots of them, and getting the controls to be able to handle all of the inputs and then deciphering what to do with the signals.

What would have happened if the dislodged part had fallen harmlessly to the ground instead of landing in the queue? I assume the brakes would have been active and stopped the train, but would the control system have essentially lost track of where it was?

Most likely you'd have a sensor fault, and the trains would just hold position where they were until the sensor operation and blocks were verified and the fault was cleared. Then they'd get another sensor fault when they tried to move that train, and they'd figure out that the problem was with the train and not the sensor, so they'd unload and remove that train for inspection.

Shades said:

The sensor by itself is inexpensive, simple, and reliable. The complexity comes in with having lots and lots of them, and getting the controls to be able to handle all of the inputs and then deciphering what to do with the signals.

Absolutely this. Once the integrators have the ride up and running, it's relatively self-sufficient other than regular maintenance. And if they've done their job well, troubleshooting problems is pretty straightforward and can be done entirely from an HMI. Either way, locating a fault in a PLC isn't an overly complex process and can be remedied fairly quickly. This is what I do on a daily basis -- but with making ranges and stovetops.

As kpjb said, if the flag had fallen off harmlessly, a fault would have been thrown. Brakes are always in a normally-closed position and require power to open or retract, so the train would stop without incident. The correlation between what the PLC expects to see from inputs, what it receives, and how it faults is once again entirely dependent on how the integrators programmed the PLC. However, I'm reasonably certain the tire drives would never kick on unless it saw the train in the block and the once in front open. My guess is that maintenance would have checked for a tripped breaker or dead power supply on the circuit of sensors, and finding that working properly would deduce the problem was with the train. A manual mode would allow them to move the trains forward and unload.

Absimilliard said:
Every time you have a ride with multiple trains, especially one that requires to move 2 or 3 trains to move at the same time in the same blocks, you need a lot of sensors to know where they are.

Yeah, I get that, but your follow up to that makes my point: You don't need to do that to get the most out of the ride. Is it any wonder or surprise that Cheetah Hunt moves one train at a time, with a similar paired station? I don't even think that Dragster moves trains in pairs anymore, if it ever did.

My point stands, the ride didn't have to be that complex.

Without talking to someone directly in charge of CPs controls and maintenance, we won't know. If they simply wanted those blocks to work as check-in/checkout, they could (probably) adapt the system they were given to do so. Similarly, if there wasn't a reason why the sensor flags needed to be removeable, they surely would've welded them onto the chassis by now.

But the sensors remain in place, so either CP wanted that level of monitoring or they're there because of some other engineering decision. I assume that if they had bypassed a large portion of their signals, they'd have no reason to keep them installed along the track.

Still, I understand Jeff's point about the system being complex. But, the complexity of the system was not what caused this accident. It was a hardware failure through and through so the number of sensors here is really just a moot point.

Of course the complexity itself didn't cause the accident, but complex things are more likely to break. An electric car has a fraction of the parts of a gas car, and requires exponentially less maintenance.

In regards to Dragster’s hydraulic system vs Red Force LSM: The hydraulic system can generate more power, but as I understand is horribly inefficient as far as how much electricity it takes to get the train up to speed. The LSMs however, are more efficient using far less electricity to launch, but can’t generate quite as much power.

Red Force takes 2 seconds to go 0 to 100 kph, which is similar to the first half of Dragster’s launch. Red Force then takes an additional 3 seconds to reach 180 mph. The reduced power results in slower acceleration, requiring a longer launch section.

I don’t remember where I heard this, but as I recall, Cedar Point didn’t have enough room to fit a ride with a LSM launch and still get a ride 400 feet tall. So they had to go with hydraulic.

Jeff said:

I don't even think that Dragster moves trains in pairs anymore, if it ever did.

My point stands, the ride didn't have to be that complex.

Agreed. And I'm not ragging on Intamin per se, but this is the same company that thought a flume ride (STR) needed pumps in the boats to get rid of water.

I'm not exactly sure what you mean by moving in pairs, but the ride does move two trains into the station simultaneously to load, and then sends them out together to launch, with train B out of the station in a waiting area while train A launches.

But they don't move out at the same time. At least, not in the video I shot in 2009.

And now my memory is messing with me, because I could have sworn they do move out together. I'm pretty sure both trains are out of the station while train A launches.

Or maybe what happens is that train B moves into train A's former spot while train A launches?

Now I've just confused myself.

EDIT:

In this video you can see both trains moving out of the station...first 31 seconds is all you have to sit through.

Yes, they are technically in motion at the same time, but my understanding was that the original intention was for all the trains to move together bumper-to-bumper. That's opposed to how they actually do it now where they keep an entire train's-length block free between them.

Has Gemini always had to close for even the slightest rain because of the sensors? Even prior to the modern CP "sprinkles close the park" rain policy?