The toughest task I’ll accomplish this month is cleaning up my workbench. We have company coming Saturday so tomorrow will be a clean the workbench day. I know my workbench is there somewhere I just haven’t seen it for awhile. Laying track is surely tougher than cleaning my workbench, never happen!Seriously just getting my layout back operational by mid December is my goal. Growing old is the pits guys!!!I tore up my control panel earlier this year to finish the installation of my Arduino signal system and do to health problems its still screwed up. I figure I have about 80 hours of work on it before its back operational. I can only do a couple of hours at a time then its rest time. Arthritis pain dings me quickly. At least it has finally cooled off here so I won’t have to work in higher temps in the garage. This summer has been a real cooker, daily temps in the garage has been over 85° for months today it was 75°.Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps.
LastspikemikeSorry to bring this up but achieving the age 83 does mean you likely have more than enough time to build another layout
How in the hell can you post something like that?I've been around here longer than you, I've have pm'd with Mel, but I don't know him well enough to say that, even if we had been drinking adult beverages all night
Henry
COB Potomac & Northern
Shenandoah Valley
York1 RR_Mel I don’t have clue as to how I positioned the original track on my layout. I don't either. I laid a lot of flex track, and I didn't pay attention to which rail was which. I generally moved things around to fit, cut rails to fit, etc. If I build a new layout in my 70s, maybe I will do it the right way.
RR_Mel I don’t have clue as to how I positioned the original track on my layout.
I don't either. I laid a lot of flex track, and I didn't pay attention to which rail was which. I generally moved things around to fit, cut rails to fit, etc.
If I build a new layout in my 70s, maybe I will do it the right way.
I did it that way too. I'm one of those guys who takes up used rail and fiddles with it or replaces it when I slightly alter the track plan.
Like many things, acquired knowledge comes from discovered mistakes.
I've soldered a sliding inside rail to a fixed inside rail because I didn't check the alignment before hand.
Trying to bend the now one piece of track that has half of the inside rail sliding and the other fixed presents interesting challenges.
- Douglas
LastspikemikeLooking closely I also notice that both Shinohara and Walthers actually leave two spike heads missing every so often in both ends of a tie allowing you to actually spike down the flex track right onto the rail.
Correct:
Track_83Shin by Edmund, on Flickr
I suggest using a pin vice to clean out the spike hole (depending on your preference of spike) as the tie end will snap off if a spike is used that fits too snugly.
Regards, Ed
RR_MelI don’t have clue as to how I positioned the original track on my layout.
York1 John
I don’t have clue as to how I positioned the original track on my layout. I did look at the replacement track and the two pieces I can see through inspection doors are to the Atlas spec, but that occurred a few years earlier than 2003. Just thinking about laying track I think I would have the sliding rail to the inside because it’s easier to fit that way.I would say that in my case it didn’t make any difference because the spikes broke on both rails. I can’t see any difference in the spikes on either side of either rail. At 83 I seriously doubt if I will ever lay any more track of any kind on any layout but if I had known back then I would have made sure it was done to spec.Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps.
Lastspikemike RR_Mel I agree with Sheldon. I too was wondering about that. I’ve been using Atlas Flex Track since it was first released and installed it both ways on three layouts. Can’t say I’ve ever seen or been instructed which side is correct. On tight curves the ties are closer one way and wider on the opposite side. I always tried to keep it the same way. Anyone know for sure what Atlas says?Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps. I do, or more correctly what they wrote in their 2003 catalogue on page 6. "you must keep in mind that the sliding rail must always be on the inside of the curve" This instruction appears right next to a nicely curved piece of flex track... curved into a lovely smooth S bend with no tangent transition!
RR_Mel I agree with Sheldon. I too was wondering about that. I’ve been using Atlas Flex Track since it was first released and installed it both ways on three layouts. Can’t say I’ve ever seen or been instructed which side is correct. On tight curves the ties are closer one way and wider on the opposite side. I always tried to keep it the same way. Anyone know for sure what Atlas says?Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps.
I agree with Sheldon.
I too was wondering about that. I’ve been using Atlas Flex Track since it was first released and installed it both ways on three layouts. Can’t say I’ve ever seen or been instructed which side is correct. On tight curves the ties are closer one way and wider on the opposite side. I always tried to keep it the same way. Anyone know for sure what Atlas says?Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps.
I do, or more correctly what they wrote in their 2003 catalogue on page 6.
"you must keep in mind that the sliding rail must always be on the inside of the curve"
This instruction appears right next to a nicely curved piece of flex track...
curved into a lovely smooth S bend with no tangent transition!
OK, fine. But again, at 36" radius, or even 30" radius, I have never had an issue.
It has been about 4 decades since I even considered bending a piece of flex track any sharper than that.
Somewhere I have few Atlas catalogs, likely older than 2003, which was 17 years ago. It would be interesting to see what information other publications from Atlas offer on the subject.
It fact, I have the Atlas custom line layout book, I think from 1957 - I will check it out....
Sheldon
Before Sheldon jumps in, I think Atlas is being conservative when they say "must". I prefer to always have the sliding rail on the inside because it creates overlap at the end(s) that snips off easier.
Also, if you curve it into a 20 something radius, or heaven forbid something sharper, the gapped side will splay the ties far apart if its on the outside. The spacing will look way out of scale.
But for broad radius curves, these issues don't really emerge and it probably doesn't matter.
Time for an off-topic sideline:
DoughlessYeah, Wild Bill was quite the teacher and test giver. I also had the (un)fortune of having him for home room for 3 years. He was actually a good guy.
I enjoyed him. Our class was a little special to him because his daughter was in our class.
In the late 60s, the education department at Kearney State College came up with a new "hands-on" physics curriculum that GISH tried. By Christmas, Smith saw that we weren't learning anything. After the Christmas break, he taught us the entire year's regular physics class in four months.
Ok, back to the topic at hand!
Lastspikemike I note that Atlas flex track only has gaps in the tie webs under one rail. The rail that's supposed to be the outside rail has no flexing gaps in the ties. Shinohara uses the same pattern. Peco and ME elect to put gaps under both rails. Also, comments elsewhere lead me to wonder if flex track is sometimes laid under tension by bending to the marked centreline and then nailing or gluing instead of bending to a point inside the marked centreline until the released curve fits the desired line and only then nailed or glued. Any thermal effects will be magnified in that situation. Ind in livable rooms the interior temperature would normally vary no more than about ten degrees (20F in America, which sounds like so much more heat). In our Canadian climate only summer weather makes a difference and our layouts are frequently in the basement anyway which tends to track below grade temperatures rather than air temperatures. We also have AC so our house may vary from 21-25C at most which is about 8F. If you get variations approaching 20 or 30F degrees I would expect some thermal stresses on the layout. A helix would be a special case. I note that Shinohara/Walthers flex track provides moulded in nail holes in the ends of the ties rather thank down the centreline of the ties. Maybe building a helix and nailing down only a well spaced number of the inside ends of the ties would protect against the thermal expansion effects described here. The flex gaps are all on that side of the track.
I note that Atlas flex track only has gaps in the tie webs under one rail. The rail that's supposed to be the outside rail has no flexing gaps in the ties. Shinohara uses the same pattern. Peco and ME elect to put gaps under both rails.
Also, comments elsewhere lead me to wonder if flex track is sometimes laid under tension by bending to the marked centreline and then nailing or gluing instead of bending to a point inside the marked centreline until the released curve fits the desired line and only then nailed or glued. Any thermal effects will be magnified in that situation.
Ind in livable rooms the interior temperature would normally vary no more than about ten degrees (20F in America, which sounds like so much more heat). In our Canadian climate only summer weather makes a difference and our layouts are frequently in the basement anyway which tends to track below grade temperatures rather than air temperatures. We also have AC so our house may vary from 21-25C at most which is about 8F.
If you get variations approaching 20 or 30F degrees I would expect some thermal stresses on the layout.
A helix would be a special case. I note that Shinohara/Walthers flex track provides moulded in nail holes in the ends of the ties rather thank down the centreline of the ties. Maybe building a helix and nailing down only a well spaced number of the inside ends of the ties would protect against the thermal expansion effects described here. The flex gaps are all on that side of the track.
I think it depends, but its obvioulsy the hard steel rail that doesn't give, causing the weakest link of the track laying process to be exposed as that spot absorbs the force.
In my case, the track shifted laterally off of the roadbed. I think because I simply did not adhere the track well enough to the roadbed. Poor caulking job there. If I would have sufficiently secured the track, it may have found the next weakest spot and bent upwards like others and sheard off the spike heads.
York1 Mr. Smith! One of my favorites. I think many of today's students would have a hard time passing his physics tests.
Mr. Smith! One of my favorites. I think many of today's students would have a hard time passing his physics tests.
Oh wow. I had forgotten that you grew up in the same town and went to the same HS I did, just about a decade before right?
Yeah, Wild Bill was quite the teacher and test giver. I also had the (un)fortune of having him for home room for 3 years.
He was actually a good guy.
gmpullman Lastspikemike I note that Atlas flex track only has gaps in the tie webs under one rail. The rsil that's supposed to be the outside rail has no flexing gaps in the ties. Shinohara uses the same pattern. Shinohara alternates (staggers) the tie gap: Atlas_Shinohara by Edmund, on Flickr Atlas left — Shinohara (Walthers) right. Both code 83. In the interest of accuracy. Thank you, Ed
Lastspikemike I note that Atlas flex track only has gaps in the tie webs under one rail. The rsil that's supposed to be the outside rail has no flexing gaps in the ties. Shinohara uses the same pattern.
Shinohara alternates (staggers) the tie gap:
Atlas_Shinohara by Edmund, on Flickr
Atlas left — Shinohara (Walthers) right. Both code 83.
In the interest of accuracy.
Thank you, Ed
In 40 or 45 years of using Atlas flex track, I have never given any consideration to the tie gaps being on the inside or the outside of the curve, with no ill effects.
And I have never read any information published by Atlas regarding this issue.
But then again, I am usually only bending it to 36" radius or larger except for some industrial trackage.
gregc maybe humidity has a greater effect than temperature
maybe humidity has a greater effect than temperature
Simon
LastspikemikeI note that Atlas flex track only has gaps in the tie webs under one rail. The rsil that's supposed to be the outside rail has no flexing gaps in the ties. Shinohara uses the same pattern.
greg - Philadelphia & Reading / Reading
DouglasYou are so correct. I didn’t remove the soldered joiners from the code 83 rails but none of the replacement code 100 track joiners are soldered.As a college grad (EE) one would have thought I would have known better but having soldered the joiners on three pervious layouts without a problem I just didn’t give it any thought. My only excuse is had been 30 years since my college days, that’s the best I can come up with.As Ron White said “you can’t fix stupid”. Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps.
RR_Mel In my disaster the track was Atlas code 100 and the ties were loosely anchored with track nails about every 12 inches in a hidden 30” radius helix (about 33’ of track) and the flex ties placement on the cork looked normal. The rails broke the spike heads off the ties. The rails were well away from their correct position on the ties, the outside rails were pushed out to the edge of the ties.I was so upset that I didn’t even think about taking pictures.It was rather interesting in that only the hidden track was not ballasted and all the none hidden track was Atlas code 83 well anchored with ballast (about 90’). Only the unballasted track was effected.I was the new guy on the block back then in a new city and bought all of my track from the same LHS. When I went in and bought the replacement track and told the owner what happened he said “bet you soldered the joiners”. I was so upset that I didn’t even think about taking pictures. Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps.
In my disaster the track was Atlas code 100 and the ties were loosely anchored with track nails about every 12 inches in a hidden 30” radius helix (about 33’ of track) and the flex ties placement on the cork looked normal. The rails broke the spike heads off the ties. The rails were well away from their correct position on the ties, the outside rails were pushed out to the edge of the ties.I was so upset that I didn’t even think about taking pictures.It was rather interesting in that only the hidden track was not ballasted and all the none hidden track was Atlas code 83 well anchored with ballast (about 90’). Only the unballasted track was effected.I was the new guy on the block back then in a new city and bought all of my track from the same LHS. When I went in and bought the replacement track and told the owner what happened he said “bet you soldered the joiners”. I was so upset that I didn’t even think about taking pictures. Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps.
From recollection of Mr. Smith's high school physics class.
The force exerted by the expansion found the weak spot in your track. The strong ballasted part of the layout transferred the force to the weak spot. Essentially, your helix became the expansion gap.
Unless you provide a deliberate gap(s) to channel the force to, the layout will find its weak spot.
OvermodIt was my first helix (1989) in my then 38yr model railroading career.In my case the rail didn’t break all the spikes in each section of flex but seriously dinged a total of 11 sections of flex track, both rails in every section. I laid most of my track in the fall and winter months (maybe around 60°s daytime, 40°s or so over night). The garage would be 104°+ most of the summer.It took almost two years to complete my 120’ mainline and the beginning of fall the second year is when I discovered the disaster. It caught me so off guard that I actually was looking of evidence of a critter not a heat expansion problem.We weren’t newbie’s to heat but not the severe 80°+ temperature swing between winter and summer in Bakersfield. The added R30 insulation (the garage roof had maybe R10) took care of the problem. The added insulation really helped the house too, well worth the expense! Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps.
RR_MelIn my disaster the track was Atlas code 100 and the ties were loosely anchored with track nails about every 12 inches in a hidden 30” radius helix (about 33’ of track) and the flex ties placement on the cork looked normal. The rails broke the spike heads off the ties.
Perhaps it bears noting that if there are going to be differential-expansion effects from a length of metal rail overconstrained at both ends, they'll manifest most notably on the outside radius of a prolonged curve (which is of course what a helix is). The 'load' is entirely on the inner edges and tops of the little plastic tabs on the outer-radius side, and being tangential the leverage on them is immense (compare the tension pushing sideways on a clothesline between two fixed ends). If you have in fact soldered this into one long effective rail with no 'give' at the ends, retaining gauge will need much more strength than unmodified flex track will likely provide. (Note that it would also be possible for rail in track laid and soldered at high temperature to pull to the inside on cold contraction...)
I think a kind of moral here is that helices need to be built with unsoldered gaps on at least one side of joiners, and with feeders to each section to compensate for any loss in 'continuity'; if there are valid places to use dielectric grease on joiners to aid in effective electrical continuity by preventing oxidation (as with aluminum house wiring) this would be at the top of the list.
riogrande5761 ATLANTIC CENTRAL he layout was DCC, and he installed feeder drops every 6 feet. Most of those drops never got connected to the buss. The rail joints were soldered. It stands to reason that if all the joints were soldered, all those drops would indeed be redundant. I'm not sure what is over-kill, but on stretches of unsoldered joints left floating, it seems reasonable to provide drops connected to the bus at fairly frequent intervals. On a single track mainline, every six feet seems reasonable and not a great deal of work. If one solders every joint much longer intervals would probably be sufficient.
ATLANTIC CENTRAL he layout was DCC, and he installed feeder drops every 6 feet. Most of those drops never got connected to the buss. The rail joints were soldered.
It stands to reason that if all the joints were soldered, all those drops would indeed be redundant.
I'm not sure what is over-kill, but on stretches of unsoldered joints left floating, it seems reasonable to provide drops connected to the bus at fairly frequent intervals. On a single track mainline, every six feet seems reasonable and not a great deal of work. If one solders every joint much longer intervals would probably be sufficient.
Given the nature of DCC, agreed, unsoldered track needs more feeders.
While I have always soldered my rail joints, or provided jumpers around a very few "expansion joints", with DC, other modelers I know have never had issues of voltage drop or other losses even with some joints not soldered.
My longest blocks are typically 50-60 feet, fed with #18 feeders that may sometimes be 20' from the #12 throttle buss that distributes the 8 throttles to the local relay panels of my Advanced Cab Control system.
For the record, track warped on my previous layout one summer. The layout was in my basement (in Canada), where there is no air conditioning. I don't think the temperature was higher than 85 degrees... Nothing major, but I noticed it when one of my steam loco's hot frame shorted on the curves (from the cowcatcher). That cowcatcher was filed down, but the warp was very much noticeable when you looked up close. The layout was pretty much a loop on a 4X6. My new layout has a few expansion gaps...
LastspikemikeIt is probable that the thermal expansion or contraction of the plastic ties is greater than that of the nickel silver rail. There's more of it to start with. As they change length together the difference must be tiny.
Interestingly, a number of the reported 'buckling' incidents refer to vertical failure, where the rails or track 'bow' vertically instead of 'sun kinking' sideways. In the 1920s some German research into 'sprung track' construction took up the issue of sun kinking, and one of their early experiments involved resistance heating the rails in order to mimick environmental heating. What they observed was just this sort of vertical bowing of the test section, rather than lateral motion; whether it was due to the rate of heating I do not now remember. It would certainly not occur if the ties were well bonded down, or substantially heavier in mass.
ATLANTIC CENTRALhe layout was DCC, and he installed feeder drops every 6 feet. Most of those drops never got connected to the buss. The rail joints were soldered.
Rio Grande. The Action Road - Focus 1977-1983
Lastspikemike Technically, the track movement due to temperature changes would be relative. To estimate this one uses the difference in the coefficient of expansion and some very small numbers are involved. It is probable that the thermal expansion or contraction of the plastic ties is greater than that of the nickel silver rail. There's more of it to start with. As they change length together the difference must be tiny. Unless your railroad is subjected to wide range of temperature change, as in temporary unheated storage or while moving house for example, I expect the differential expansion or contraction between the plastic ties and the rails would be insignificant. The whole layout responds to temperature, unlike the prototype. Rail distortion should be very rare either way: soldered or not, nailed down or glued.
Technically, the track movement due to temperature changes would be relative.
To estimate this one uses the difference in the coefficient of expansion and some very small numbers are involved.
It is probable that the thermal expansion or contraction of the plastic ties is greater than that of the nickel silver rail. There's more of it to start with. As they change length together the difference must be tiny.
Unless your railroad is subjected to wide range of temperature change, as in temporary unheated storage or while moving house for example, I expect the differential expansion or contraction between the plastic ties and the rails would be insignificant.
The whole layout responds to temperature, unlike the prototype. Rail distortion should be very rare either way: soldered or not, nailed down or glued.
Yes Mike, good points. Previous discussions on this forum concluded that there is very little expansion of the actual track materials as you mention.
Most expansion and contraction comes from the wood we use in our benchwork, which then carries to the track as we ultimately attach our track to the benchwork. The layout becomes fixed layers of different materials with different properties, and most are at the mercy of the more violent expansion/contraction rate of the wood benchwork.
If you run a layout in its temporary stage, like I do without permanently attaching the track to the benchwork as I test-build, the track doesn't move with the expanding/contracting benchwork. Its only after its attached does the surprise happen if the track is not gapped at some point.
Dehumidifiers would help.
Building benchwork out of metal would help.
I can think of a couple of forum members who did that because of climate reasons.
Doughless ATLANTIC CENTRAL Like Wayne I'm a DC operator. I have been soldering all my rail joints within each electrical control section (commonly called blocks) for 50 years now. Some blocks are 50-60 feet long. Because all my block feeders on the mainline run thru inductive dectectors for the signal system, installing feeders every 3-6 feet like many people do with DCC would be very impractical. My layouts have always been in climate controlled stable environments and expansion/contraction has never been an issue. I say solder all or nearly all, and possibly make some allowance for expansion. You can allow some expansion joints and still solder flexible jumpers around those rail joiners. Shedlon Let me add some points. I run DCC for onboard sound purposes only, and I'm never concerned about electrical current petering out because of excessive length of track. It simply never mattered. While some may solder a feeder to every piece of track, I think that is a bit conservative. I do solder the turnouts. My biggest concern is turnout failure later, so I have a feeder to the tracks that leads to each leg of the turnout, but that feeder might be 10 feet away. Because I build mainly switching layouts, there is always a turnout close by. So that type of layout forces more feeders. If I had a 30 foot run with no turnouts, I would probably just have a pair of feeders at each end of the run (and a small gap in the middle). In an undehumidified basment in Indiana, there would be bigger differences in the humidity levels between summer and winter, when dry Canadian air would dip into the midwest. I did notice buckling along the 35 foot long un-gapped section one spring,but cutting a small 1/16th of an inch gap cured the issue. No need to go over board on the expansion gapping. Obviously there is a need for a feeder on each side of each gap, but how far away the feeder is from the gap is not a huge issue. Could be 20 feet, IMO.
ATLANTIC CENTRAL Like Wayne I'm a DC operator. I have been soldering all my rail joints within each electrical control section (commonly called blocks) for 50 years now. Some blocks are 50-60 feet long. Because all my block feeders on the mainline run thru inductive dectectors for the signal system, installing feeders every 3-6 feet like many people do with DCC would be very impractical. My layouts have always been in climate controlled stable environments and expansion/contraction has never been an issue. I say solder all or nearly all, and possibly make some allowance for expansion. You can allow some expansion joints and still solder flexible jumpers around those rail joiners. Shedlon
Like Wayne I'm a DC operator. I have been soldering all my rail joints within each electrical control section (commonly called blocks) for 50 years now.
Some blocks are 50-60 feet long.
Because all my block feeders on the mainline run thru inductive dectectors for the signal system, installing feeders every 3-6 feet like many people do with DCC would be very impractical.
My layouts have always been in climate controlled stable environments and expansion/contraction has never been an issue.
I say solder all or nearly all, and possibly make some allowance for expansion.
You can allow some expansion joints and still solder flexible jumpers around those rail joiners.
Shedlon
Let me add some points.
I run DCC for onboard sound purposes only, and I'm never concerned about electrical current petering out because of excessive length of track. It simply never mattered. While some may solder a feeder to every piece of track, I think that is a bit conservative.
I do solder the turnouts. My biggest concern is turnout failure later, so I have a feeder to the tracks that leads to each leg of the turnout, but that feeder might be 10 feet away.
Because I build mainly switching layouts, there is always a turnout close by. So that type of layout forces more feeders.
If I had a 30 foot run with no turnouts, I would probably just have a pair of feeders at each end of the run (and a small gap in the middle).
In an undehumidified basment in Indiana, there would be bigger differences in the humidity levels between summer and winter, when dry Canadian air would dip into the midwest. I did notice buckling along the 35 foot long un-gapped section one spring,but cutting a small 1/16th of an inch gap cured the issue. No need to go over board on the expansion gapping. Obviously there is a need for a feeder on each side of each gap, but how far away the feeder is from the gap is not a huge issue. Could be 20 feet, IMO.
Great points, completely agreed.
I too feel the every 3-6 feet feeder thing is way overkill. I have been involved in the building of, or DCC conversion of a number of large layouts, 1500 sq ft and larger.
One layout I designed and helped build for a friend had two double track helices down to a staging level. The layout was DCC, and he installed feeder drops every 6 feet. Most of those drops never got connected to the buss. The rail joints were soldered. The layout had a double track mainline over 600 feet long. It ran flawlessly for over a dozen years with only about 6-8 feeders on each loop until he dismantled it to move.
Just one example.
As a DC operator, turnouts are mostly near block boundaries and on the mainline represent a kind of sub-block called an "x section", so they always have separate feeders nearby.
I had a large temperature swing in my garage when we first moved to Bakersfield (1987) and that’s where I built my layout. We had never lived in a location that had large differences in temperature so like all my other layouts over the years I soldered all the joiners. That was a bad mistake. On the second year I lost 11 sections of flex track do to the large temperature swing.Looked a lot like this. (Bakersfield Earthquake 1952)
The track was Atlas code 100 Flex and one or both rails broke loose from the tie spikes.The track couldn’t take the 70°+ temperature swing (35° to 110° in the garage) from winter to summer. We had the garage insolated (R40) and no more problems but then I didn’t solder all the joiners when I replaced the track either. I should have known better and left room for expansion.Mel My Model Railroad http://melvineperry.blogspot.com/ Bakersfield, California I'm beginning to realize that aging is not for wimps.
I run DCC for onboard sound purposes only, and I'm never concerned about electrical current petering out because of excessive length of track. While some may solder a feeder to every piece of track, I think that is conservative.
Because I build mainly switching layouts, there is always a turnout close by. That type of layout forces more feeders, so I can rarely get a long run of track without the need for a feeder for potential turnout failure reasons.