LimitState:RING - Experimental Validation

Download
Get LimitState:RING

The solutions provided by LimitState:RING have been validated against the results from a number of full- and small-scale experimental bridge tests:

Full Scale

Bolton laboratory tests
TRL field tests

Small Scale

Sheffield laboratory tests
Salford laboratory tests

Bolton laboratory tests (full-scale)

In Bolton, UK, in the early 1990's a number of 3m and 5m span masonry arch bridges were tested in the laboratory. A key advantage of these tests over field tests was that the internal constructional details and material properties were known.

LimitState:RING was originally developed to assist with the interpretation of the results from these laboratory tests.

In Table 1, sample LimitState:RING results are presented alongside the experimental test results. It is clear from Table 1 that predictions are quite conservative when the default soil angle of friction is used (A), but become much more realistic when the measured value is used (B).

Bridge^#	Description	Experimental collapse load (kN)	LimitState:RING Analysis (kN)		LimitState:RING (Case B) / Experimental
			(A)	(B)
			Default soil properties (measured unit weight)	As (A) but using measured angle of soil friction (60^o)
3-1	3m single span	540	265	442	82%
5-1	5m single span	1720^*	1178	1915	111%^*
5-2	5m single span (debonded arch rings)	500	253	400	80%
Multi-2	3m triple span	320	202	320	100%
^* The experimental collapse load of this bridge was reduced by the sudden onset of partial ring separation. # Only bridges with detached spandrel walls are included since these behave in a two dimensional manner.

Table 1 - Results from LimitState:RING analysis of Bolton masonry arch bridge tests

TRL field tests (full-scale)

Bridgemill TRL arch bridge experiment In the late 1980s and early 1990s, the Transport and Road Research Laboratory (now TRL) carried out a series of load tests to collapse on redundant arch bridges.

Most bridges failed with four hinge mechanisms, although some were reported as failing by ‘three hinge snap through’ or in ‘compression’ (material failure). It was likely that many were restrained considerably by their attached spandrel walls and/or masonry backing.

In 2001, TRL were commissioned to independently validate an early version of LimitState:RING and other available masonry arch bridge analysis software. The programs were used to predict the carrying capacities of 5 of the bridges tested in Bolton more than a decade previously. Details taken from the TRL report relating to LimitState:RING for 4 of the bridges are provided in Table 2.

Agreement between the LimitState:RING predictions and the full-scale test results was good and, based on this evidence, Network Rail have confirmed that LimitState:RING is a suitable program for use to assess masonry arch bridges on the UK rail network.

Bridge Name	Collapse Load (Theoretical / Experimental)
Torksey	81%
Bridgemill	100%
Barlae	92%
Preston	90%

Table 2 - Results from LimitState:RING analysis of TRL masonry arch bridge tests

Sheffield laboratory tests (small-scale)

Small-scale arch bridge experiment A series of small-scale tests were performed at the University of Sheffield to confirm the relative importance of:

Passive restraint effects (i.e. as parts of the arch barrel move into the fill), and
Live load dispersion effects (i.e. as the live load spreads through the fill).

The LimitState:RING and experimental results are compared in Table 3. It is evident that the LimitState:RING predictions are remarkably good (all within 10% of the experimental results), verifying that the LimitState:RING soil model is capable of capturing the key effects of backfill.

Test Name [key^#]	Experimental peak load capacity (N) [Without extended keystone]			LimitState:RING calculated peak load capacity (N)	LimitState:RING Result / Mean Experimental Result
T1 [---]	107 [104]	107 [104]	107 [104]	99	93%
T2 [-P-]	141	142	140	133	94%
T3 [AP-]	107 [104]	107 [104]	107 [104]	132	96%
T4 [APL]	107 [104]	107 [104]	107 [104]	187	104%
T5 [A--]	103	104	100	97	95%
T6 [A-L]	130	131	136	136	103%
# Factors considered: A = Active; P = Passive; L = Load Spreading

Table 3 - Results from LimitState:RING analysis of Sheffield small-scale masonry arch bridge tests

Salford laboratory tests (full-scale)

To better establish the nature of the soil-arch interaction which takes place, a series of bridges were tested at the University of Salford, UK.

Experimental and analysis results are provided in Table 4, where it is clear that, when the measured soil strength parameters are used (B), LimitState:RING provides a very accurate prediction of the load carrying capacity:

Bridge	Description	Experimental collapse load (kN)	LimitState:RING Analysis (kN)		LimitState:RING (Case B) / Experimental
			(A)	(B)
			Default soil properties (measured unit weight)	As (A) but using measured angle of soil friction (60^o)
1	3m single span (limestone fill)	126	84	122	96.5%
2	3m single span (clay fill)	92	88	95.5	104%

Table 4 - Results from LimitState:RING analysis of Salford masonry arch bridge tests

References

Bolton tests

Gilbert, M. and Melbourne, C. 1994 Rigid-block analysis of masonry structures. The Structural Engineer 72, 356–360

Melbourne, C. and Gilbert, M. 1995 The behaviour of multiring brickwork arch bridges. The Structural Engineer 73, 39–47

Melbourne, C., Gilbert, M. & Wagstaff, M. 1997 The collapse behaviour of multi-span brickwork arch bridges. The Structural Engineer 75(17), 297–305

TRL tests

Page, J. 1993 Masonry Arch Bridges. HMSO, UK.

Macfarlane, A. and Ricketts, N. 2001 Evaluation of existing software for the assessment of masonry arch bridges, Technical report, TRL report to Railtrack.

Sheffield tests

Gilbert, M., Smith, C., Wang, J., Callaway, P. and Melbourne, C. 2007 Small and large-scale experimental studies of soil-arch interaction in masonry bridges, Proc. 5th International Arch Bridges Conference, Madeira, 381–388.

Learn more