Metro to take 7000 series out of service after defect was discovered in axle

The D.C. Metrorail will pull almost 60% of its rail fleet from service starting Monday, Oct. 18, the agency announced Sunday.

The agency is removing its 7000-series railcars from service after the National Transportation Safety Board found defects in several of them while investigating the derailment of a blue line railcar last week, on Oct. 12.

https://www.msn.com/en-us/news/us/60-of ... d=msedgntp

https://wmata.com/about/news/Metrorail- ... ber-18.cfm

Good luck to anyone braving the announced 30-minute headways on all lines today.

30 mins is generous. I was stuck at Reston for closer to an hour than 30mins. Platform was pretty much full by the time the train pulled up.

To put it in context, we're still running about 28% of ridership measured from 2019. So it's bad, but not too bad. WMATA can get away with it, but I'd prep some bus bridges.

Well, this is not looking good. Breaking news per https://www.nbcwashington.com/news/loca ... b/2836752/

-the train that de-railed had done so 2x already that day, but both times self-corrected

-Metro has known about this defect since 2017

"wheel gauge anomalies"

WMATA Safey [sic] Commission order

https://wmsc.gov/wp-content/uploads/202 ... ervice.pdf

I blame that article on governing.com touting how Metro had rebounded.

Some additional details in a longer WaPo article - The chair of the National Transportatio ... re unsafe.

Among them, "wheel assemblies on Metro rail cars at fault in last week’s Blue Line derailment had failed 31 times since 2017 — and that renewed inspections last week identified almost two dozen more cars with assemblies that were unsafe." In addition, "data gathered by NTSB investigators indicated problems with rail cars were getting progressively worse: Metro identified two cars with failures in 2017, two more in 2018, four in 2019 and five in 2020. Before the derailment, the agency had found 18 unsafe cars this year, and inspections last week turned up 21 more. Homendy said about 200 still need to be inspected."

Finally, as to the derailment, "wheels on the car involved in last week’s derailment had shifted outward, so when the car passed over a switch, it bumped the car off the rails. Gordon said passing another switch knocked it back on, explaining the two previous derailments. But at Rosslyn, where the train left the tracks for the final time, Gordon said there was not another switch."

The story mentions that Metro has been working with Kawasaki on this issue, but does not have details.

So this train was dragging a wheelset on the ties for a whole distance between switches? I guess if it popped off within a crossover, the distance is short, but that must have been a HECK of a ride for anyone in that car.

NTSB news conference today

https://youtu.be/UOWAE1kNxUQ

What I see as the #1 question:

If an axle set passes inspection, how long can it be safely run until the next inspection?
Clearly 90 days is too long...

farecard
What I see as the #1 question:

If an axle set passes inspection, how long can it be safely run until the next inspection?
Clearly 90 days is too long...

There should be no change in the distance between the inside wheel faces over 90 days.

I Saw the NTSB Media Briefing.

This is sounding like an issue with the assembly of the wheel sets. Seems like the vendor that made the wheel sets for Kawasaki were fitting wheels to the axles that had interference fits that were out of tolerance. Either the inside diameter of the wheel bore was to big or the outside diameter of the axle was to small or combination of both. It could also be that the interference fit tolerances used were not correct in the first place.

The tolerance used in interference fits are measured in ten thousands of an inch or one hundreds of mileometer. The temperature of two parts being interference fitted must be same when the measurements are made.

Sand Box John wrote: ↑Tue Oct 19, 2021 1:46 am farecard
What I see as the #1 question:

If an axle set passes inspection, how long can it be safely run until the next inspection?
Clearly 90 days is too long...

There should be no change in the distance between the inside wheel faces over 90 days.

I Saw the NTSB Media Briefing.

This is sounding like an issue with the assembly of the wheel sets. Seems like the vendor that made the wheel sets for Kawasaki were fitting wheels to the axles that had interference fits that were out of tolerance. Either the inside diameter of the wheel bore was to big or the outside diameter of the axle was to small or combination of both. It could also be that the interference fit tolerances used were not correct in the first place.

The tolerance used in interference fits are measured in ten thousands of an inch or one hundreds of mileometer. The temperature of two parts being interference fitted must be same when the measurements are made.

I'm not too familiar with the 7000 series, but they look like inside bearing trucks, so the wheels are "pressed" on to a shoulder that sets the wheel spacing. If the wheel is loose, it can only go outwards. Agree with Sand Box John here - this is a major design or manufacturing defect. If it's affecting so many axles, I'm inclined to believe it's a design issue.
Once an interference fit slips, usually both parts are ruined. These cars may not be back in service until they can get new wheels and axles installed. Might be possible to machine the axle to clean up the surface, apply a surface weld to the bore of the wheel, then machine the weld back to proper tolerance, and shrink fit the pair back together. But feasibility of that repair depends on lots of details of the design, material, and condition of the parts.

so the rail guage on our system is 1429in and not the stand 1435in, could that be at play here?

YOLO wrote: ↑Tue Oct 19, 2021 11:16 am so the rail guage on our system is 1429in and not the stand 1435in, could that be at play here?

Doubtful. Changing the length a few percent is hardly rocket science.