**5. Conclusions**

Bed degradation is a common river process and can cause general scour depth at instream infrastructures, leading to structural damages or failures. Grade control structures (GCSs) are conventional countermeasures for the general scour at instream structures. However, improper scour design for the GCS can destabilize both the GCS and the upstream river reach. This study conducts a 1:80 scaled model test based on a 1.3 km-long reach of Shiting River, China, investigating the effects of GCS failure (drop in GCD) on the upstream river bed morphology.

The experimental results indicate that, for each GCD *z*, the upstream river bed elevation decreases with increasing discharge *Q*. For a lower *z*, the difference caused by increasing *Q* is greater. For *z* = 533 m and *z* = 527 m, a head cut erosion area exists upstream of the GCS for *P* ≥ 20%. For *P* ≤ 10%, the head cut erosion area of *z* = 533 m and *z* = 527 m vanishes. The average upstream bed gradient *i* is approximately independent of *Q* for *z* = 539 m; *i* decreases with increasing *Q* for *z* = 533 m and *z* = 527 m. The experimental results also indicate that, for each *z*, the increase in *Q* can widen and deepen the channel. The difference in the channel width and depth caused by increasing *Q* is greater for a lower *z*. For each *Q*, the decrease in *z* induces a wider and deeper channel.

The maximum general scour depths *dp* at the 105th Provincial Highway Bridge are analyzed. For *z* = 539 m, *dp* is approximately independent of *Q*. For a lower *z*, *dp* increases slightly with increasing

*Q*. For the new GCS design plan (*z* = 533 m), the general scour depth at the bridge site is still significant (*dp* = 9.4–10.9 m). Thus, this study suggests building a GCS with *z* ≥ 539 m or a new bridge with much deeper foundations.

**Author Contributions:** X.M. did all the experiments, data process and part of result analysis. L.W. wrote the paper and did most of the results analysis. R.N., K.Y., X.L. proofread the paper, and provided a lot of valuable advice on experimental operation, result presentation and writing.

**Funding:** This research is supported by the National Key Research and Development Program of China (2016YFC0402302), the National Natural Scientific Foundation of China (51809187) and the Fundamental Research Funds for the Central Universities (YJ201935).

**Acknowledgments:** The authors would like to thank the editors of the Special Issue of "Experimental, Numerical and Field Approaches to Scour Research" for their invitations.

**Conflicts of Interest:** The authors declare no conflict of interest.
