Fillet Depth [SIG]
By taking girder deflection into account based on the concrete deck casting (for both pouring and hardening stages), the fillet depth may need to be recalculated. To achieve this, this object can be utilized. However, for the deflections based on DC1 (non-composite dead load case) and DC2 (composite dead load case) to be calculated successfully, the analysis and design must be completed.
General
DC1: Specify the non-composite dead load case using this parameter.
DC2: Specify the composite dead load case using this parameter.
Girder: Specify the related girder.
What the report shows
The fillet is read from the girder as modelled: the concrete from the deck bottom to the top of the top flange at every station. That is the Haunch Thickness minus the top flange thickness, plus two corrections for what the girder line does not follow between its supports. The girder takes the deck thickness and the cross slope at its supports and splice ends and runs straight between them, while the deck soffit follows the actual deck thickness and the actual cross slope. So extra deck thickness between supports reduces the fillet, and in a superelevation transition the fillet changes by the difference between the actual cross slope at the girder and the straight-line one, which can be several inches. The table lists, per station, the profile grade line and its chord between supports, the DC1 and DC2 deflections, the fillet as modelled, the fillet after those deflections (a downward deflection deepens it), the top-flange splice plate thickness wherever the station falls on a field splice's outer top plate, and the clear fillet above the plate.
There is no minimum-fillet input on this object any more. To set the haunch so that the fillet stays within limits, use Compute Linear Haunch on the girder's Haunch Thickness column. It asks for the minimum and maximum fillet depth and for one way to straighten the girder while the deck follows the vertical curve: between every pair of supports (the default), over every span that holds part of a sag curve, from support to support, or over station ranges you enter, each Straight or Follow. A fabricated girder piece may be straight or cambered but never sagging, so the sag-curve choice straightens whole spans and the girder bends only at supports. Wherever it is not straightened, the girder follows the alignment and keeps the profile's own camber: its haunch is straight between supports. Two kinds of stretch beside a straightened part may instead take a constant haunch: from a part end inside a span or in an overhang, and over an overhang beyond a straightened part. Compute tries each such stretch both ways (every combination for up to five of them; with more, a search starting from every such stretch varying linearly and from every one constant) and keeps the girder whose worst sag is least, then the next worst, and so on, counting sags within 1/32 in (0.8 mm) of each other as equal but never a sag as equal to none; then the one exceeding the maximum least; then the one using the least concrete. A constant stretch runs from the part end to the nearest interior support, or to the girder end where there is none between; between the supports inside a following part the haunch stays straight. A span next to a straightened part ramps down to the minimum; a span between two straightened parts takes their haunches at its ends. The maximum applies everywhere and is reported where it cannot be kept. A girder counts as not sagging within 1/16 in (1.5 mm). The sag-curve choice straightens every span the profile would make sag by more, and where a sag sits on an end bearing or wholly in an overhang it straightens the end span together with its overhang, so the girder does not turn up there. A sag curve shorter than 2 ft (0.6 m), a sag grade break among them, or one that reaches less than 1 ft (0.3 m) into a span straightens that span only if the span would otherwise sag more than 1/16 in (1.5 mm). With station ranges Compute warns about any span left sagging below the straight line between its supports, or an overhang turned up above the line of its span. With both limits blank it does not write a girder whose top flange would rise above the deck bottom; where the flange is already set into the deck at a span's own supports, only going deeper than that counts, and an overhang beside such a support may go as much deeper as the straight girder carries it past its bearing (the profile's own drop and the slope between today's support haunches), which only limits can change. It does not apply to a deck carried by transverse beams. It clears the top-flange splice plates. The girder's read-only Min./Max. Fillet Depth columns then show the result, measured the same way as this report.
The report's final section plots, for each girder, the top of deck and the top of the top flange along the girder, and the fillet depth as modelled and after the DC1+DC2 deflection, with the least and greatest fillet marked.
Last updated